{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1670,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005988023952095808,
      "grad_norm": 5.398237705230713,
      "learning_rate": 1.1976047904191619e-06,
      "loss": 5.2288,
      "step": 1
    },
    {
      "epoch": 0.0029940119760479044,
      "grad_norm": 4.768432140350342,
      "learning_rate": 5.9880239520958085e-06,
      "loss": 5.1682,
      "step": 5
    },
    {
      "epoch": 0.005988023952095809,
      "grad_norm": 4.144229412078857,
      "learning_rate": 1.1976047904191617e-05,
      "loss": 5.1486,
      "step": 10
    },
    {
      "epoch": 0.008982035928143712,
      "grad_norm": 3.3602042198181152,
      "learning_rate": 1.7964071856287426e-05,
      "loss": 5.0451,
      "step": 15
    },
    {
      "epoch": 0.011976047904191617,
      "grad_norm": 3.0493969917297363,
      "learning_rate": 2.3952095808383234e-05,
      "loss": 4.9569,
      "step": 20
    },
    {
      "epoch": 0.014970059880239521,
      "grad_norm": 2.564634323120117,
      "learning_rate": 2.994011976047904e-05,
      "loss": 4.8593,
      "step": 25
    },
    {
      "epoch": 0.017964071856287425,
      "grad_norm": 2.1028244495391846,
      "learning_rate": 3.592814371257485e-05,
      "loss": 4.7325,
      "step": 30
    },
    {
      "epoch": 0.020958083832335328,
      "grad_norm": 2.0943355560302734,
      "learning_rate": 4.191616766467066e-05,
      "loss": 4.6032,
      "step": 35
    },
    {
      "epoch": 0.023952095808383235,
      "grad_norm": 2.1831929683685303,
      "learning_rate": 4.790419161676647e-05,
      "loss": 4.5291,
      "step": 40
    },
    {
      "epoch": 0.02694610778443114,
      "grad_norm": 2.2486143112182617,
      "learning_rate": 5.389221556886228e-05,
      "loss": 4.4485,
      "step": 45
    },
    {
      "epoch": 0.029940119760479042,
      "grad_norm": 1.977081537246704,
      "learning_rate": 5.988023952095808e-05,
      "loss": 4.3569,
      "step": 50
    },
    {
      "epoch": 0.03293413173652695,
      "grad_norm": 2.1440491676330566,
      "learning_rate": 6.58682634730539e-05,
      "loss": 4.2673,
      "step": 55
    },
    {
      "epoch": 0.03592814371257485,
      "grad_norm": 1.845775842666626,
      "learning_rate": 7.18562874251497e-05,
      "loss": 4.1614,
      "step": 60
    },
    {
      "epoch": 0.038922155688622756,
      "grad_norm": 2.0755035877227783,
      "learning_rate": 7.784431137724552e-05,
      "loss": 4.1784,
      "step": 65
    },
    {
      "epoch": 0.041916167664670656,
      "grad_norm": 2.0788209438323975,
      "learning_rate": 8.383233532934131e-05,
      "loss": 4.0587,
      "step": 70
    },
    {
      "epoch": 0.04491017964071856,
      "grad_norm": 2.0177366733551025,
      "learning_rate": 8.982035928143712e-05,
      "loss": 4.0414,
      "step": 75
    },
    {
      "epoch": 0.04790419161676647,
      "grad_norm": 1.9424874782562256,
      "learning_rate": 9.580838323353294e-05,
      "loss": 3.9817,
      "step": 80
    },
    {
      "epoch": 0.05089820359281437,
      "grad_norm": 2.1245999336242676,
      "learning_rate": 0.00010179640718562875,
      "loss": 3.8902,
      "step": 85
    },
    {
      "epoch": 0.05389221556886228,
      "grad_norm": 1.7750043869018555,
      "learning_rate": 0.00010778443113772456,
      "loss": 3.8991,
      "step": 90
    },
    {
      "epoch": 0.05688622754491018,
      "grad_norm": 1.888543725013733,
      "learning_rate": 0.00011377245508982037,
      "loss": 3.8088,
      "step": 95
    },
    {
      "epoch": 0.059880239520958084,
      "grad_norm": 1.696926236152649,
      "learning_rate": 0.00011976047904191617,
      "loss": 3.814,
      "step": 100
    },
    {
      "epoch": 0.06287425149700598,
      "grad_norm": 1.7164498567581177,
      "learning_rate": 0.00012574850299401196,
      "loss": 3.7896,
      "step": 105
    },
    {
      "epoch": 0.0658682634730539,
      "grad_norm": 1.924892544746399,
      "learning_rate": 0.0001317365269461078,
      "loss": 3.733,
      "step": 110
    },
    {
      "epoch": 0.0688622754491018,
      "grad_norm": 2.0697178840637207,
      "learning_rate": 0.00013772455089820359,
      "loss": 3.7365,
      "step": 115
    },
    {
      "epoch": 0.0718562874251497,
      "grad_norm": 1.8498303890228271,
      "learning_rate": 0.0001437125748502994,
      "loss": 3.6889,
      "step": 120
    },
    {
      "epoch": 0.0748502994011976,
      "grad_norm": 1.799263596534729,
      "learning_rate": 0.0001497005988023952,
      "loss": 3.6785,
      "step": 125
    },
    {
      "epoch": 0.07784431137724551,
      "grad_norm": 1.8041021823883057,
      "learning_rate": 0.00015568862275449103,
      "loss": 3.6193,
      "step": 130
    },
    {
      "epoch": 0.08083832335329341,
      "grad_norm": 2.032904863357544,
      "learning_rate": 0.00016167664670658683,
      "loss": 3.5986,
      "step": 135
    },
    {
      "epoch": 0.08383233532934131,
      "grad_norm": 2.044524908065796,
      "learning_rate": 0.00016766467065868263,
      "loss": 3.549,
      "step": 140
    },
    {
      "epoch": 0.08682634730538923,
      "grad_norm": 2.168320655822754,
      "learning_rate": 0.00017365269461077845,
      "loss": 3.5249,
      "step": 145
    },
    {
      "epoch": 0.08982035928143713,
      "grad_norm": 1.8781085014343262,
      "learning_rate": 0.00017964071856287425,
      "loss": 3.5475,
      "step": 150
    },
    {
      "epoch": 0.09281437125748503,
      "grad_norm": 1.6960855722427368,
      "learning_rate": 0.00018562874251497007,
      "loss": 3.5204,
      "step": 155
    },
    {
      "epoch": 0.09580838323353294,
      "grad_norm": 1.9243298768997192,
      "learning_rate": 0.00019161676646706587,
      "loss": 3.5206,
      "step": 160
    },
    {
      "epoch": 0.09880239520958084,
      "grad_norm": 2.000701904296875,
      "learning_rate": 0.0001976047904191617,
      "loss": 3.5163,
      "step": 165
    },
    {
      "epoch": 0.10179640718562874,
      "grad_norm": 1.6321606636047363,
      "learning_rate": 0.00019999803395762152,
      "loss": 3.4382,
      "step": 170
    },
    {
      "epoch": 0.10479041916167664,
      "grad_norm": 1.5973315238952637,
      "learning_rate": 0.00019998601953415373,
      "loss": 3.4415,
      "step": 175
    },
    {
      "epoch": 0.10778443113772455,
      "grad_norm": 1.6465246677398682,
      "learning_rate": 0.00019996308424365594,
      "loss": 3.4091,
      "step": 180
    },
    {
      "epoch": 0.11077844311377245,
      "grad_norm": 1.6946362257003784,
      "learning_rate": 0.00019992923059121106,
      "loss": 3.3752,
      "step": 185
    },
    {
      "epoch": 0.11377245508982035,
      "grad_norm": 1.602065086364746,
      "learning_rate": 0.0001998844622744483,
      "loss": 3.415,
      "step": 190
    },
    {
      "epoch": 0.11676646706586827,
      "grad_norm": 1.8944158554077148,
      "learning_rate": 0.0001998287841831396,
      "loss": 3.3535,
      "step": 195
    },
    {
      "epoch": 0.11976047904191617,
      "grad_norm": 2.0829012393951416,
      "learning_rate": 0.00019976220239866562,
      "loss": 3.3339,
      "step": 200
    },
    {
      "epoch": 0.12275449101796407,
      "grad_norm": 1.8381649255752563,
      "learning_rate": 0.00019968472419335106,
      "loss": 3.3609,
      "step": 205
    },
    {
      "epoch": 0.12574850299401197,
      "grad_norm": 1.864918828010559,
      "learning_rate": 0.00019959635802967087,
      "loss": 3.3285,
      "step": 210
    },
    {
      "epoch": 0.12874251497005987,
      "grad_norm": 2.0552523136138916,
      "learning_rate": 0.00019949711355932566,
      "loss": 3.3141,
      "step": 215
    },
    {
      "epoch": 0.1317365269461078,
      "grad_norm": 1.4688948392868042,
      "learning_rate": 0.0001993870016221875,
      "loss": 3.3043,
      "step": 220
    },
    {
      "epoch": 0.1347305389221557,
      "grad_norm": 1.7277065515518188,
      "learning_rate": 0.000199266034245116,
      "loss": 3.3307,
      "step": 225
    },
    {
      "epoch": 0.1377245508982036,
      "grad_norm": 1.6926337480545044,
      "learning_rate": 0.0001991342246406448,
      "loss": 3.302,
      "step": 230
    },
    {
      "epoch": 0.1407185628742515,
      "grad_norm": 1.7383888959884644,
      "learning_rate": 0.00019899158720553824,
      "loss": 3.2949,
      "step": 235
    },
    {
      "epoch": 0.1437125748502994,
      "grad_norm": 1.4267903566360474,
      "learning_rate": 0.00019883813751921903,
      "loss": 3.2932,
      "step": 240
    },
    {
      "epoch": 0.1467065868263473,
      "grad_norm": 1.4597861766815186,
      "learning_rate": 0.00019867389234206654,
      "loss": 3.2629,
      "step": 245
    },
    {
      "epoch": 0.1497005988023952,
      "grad_norm": 1.6646181344985962,
      "learning_rate": 0.00019849886961358621,
      "loss": 3.2334,
      "step": 250
    },
    {
      "epoch": 0.15269461077844312,
      "grad_norm": 1.46786367893219,
      "learning_rate": 0.0001983130884504501,
      "loss": 3.2368,
      "step": 255
    },
    {
      "epoch": 0.15568862275449102,
      "grad_norm": 1.3595575094223022,
      "learning_rate": 0.00019811656914440885,
      "loss": 3.2124,
      "step": 260
    },
    {
      "epoch": 0.15868263473053892,
      "grad_norm": 1.7106595039367676,
      "learning_rate": 0.0001979093331600754,
      "loss": 3.2399,
      "step": 265
    },
    {
      "epoch": 0.16167664670658682,
      "grad_norm": 1.7180885076522827,
      "learning_rate": 0.0001976914031325806,
      "loss": 3.1848,
      "step": 270
    },
    {
      "epoch": 0.16467065868263472,
      "grad_norm": 1.5397311449050903,
      "learning_rate": 0.0001974628028651007,
      "loss": 3.1918,
      "step": 275
    },
    {
      "epoch": 0.16766467065868262,
      "grad_norm": 1.2814067602157593,
      "learning_rate": 0.00019722355732625774,
      "loss": 3.1661,
      "step": 280
    },
    {
      "epoch": 0.17065868263473055,
      "grad_norm": 1.3572536706924438,
      "learning_rate": 0.0001969736926473921,
      "loss": 3.2363,
      "step": 285
    },
    {
      "epoch": 0.17365269461077845,
      "grad_norm": 1.1298632621765137,
      "learning_rate": 0.0001967132361197086,
      "loss": 3.1441,
      "step": 290
    },
    {
      "epoch": 0.17664670658682635,
      "grad_norm": 1.3632957935333252,
      "learning_rate": 0.00019644221619129548,
      "loss": 3.1402,
      "step": 295
    },
    {
      "epoch": 0.17964071856287425,
      "grad_norm": 1.4946106672286987,
      "learning_rate": 0.00019616066246401717,
      "loss": 3.1352,
      "step": 300
    },
    {
      "epoch": 0.18263473053892215,
      "grad_norm": 1.3740743398666382,
      "learning_rate": 0.00019586860569028124,
      "loss": 3.1609,
      "step": 305
    },
    {
      "epoch": 0.18562874251497005,
      "grad_norm": 1.359784483909607,
      "learning_rate": 0.0001955660777696793,
      "loss": 3.1191,
      "step": 310
    },
    {
      "epoch": 0.18862275449101795,
      "grad_norm": 1.0663578510284424,
      "learning_rate": 0.00019525311174550285,
      "loss": 3.1508,
      "step": 315
    },
    {
      "epoch": 0.19161676646706588,
      "grad_norm": 1.2853511571884155,
      "learning_rate": 0.00019492974180113426,
      "loss": 3.1197,
      "step": 320
    },
    {
      "epoch": 0.19461077844311378,
      "grad_norm": 1.2358075380325317,
      "learning_rate": 0.00019459600325631303,
      "loss": 3.0737,
      "step": 325
    },
    {
      "epoch": 0.19760479041916168,
      "grad_norm": 1.2823344469070435,
      "learning_rate": 0.0001942519325632781,
      "loss": 3.1486,
      "step": 330
    },
    {
      "epoch": 0.20059880239520958,
      "grad_norm": 1.0905367136001587,
      "learning_rate": 0.00019389756730278627,
      "loss": 3.1254,
      "step": 335
    },
    {
      "epoch": 0.20359281437125748,
      "grad_norm": 1.1666399240493774,
      "learning_rate": 0.00019353294618000758,
      "loss": 3.0982,
      "step": 340
    },
    {
      "epoch": 0.20658682634730538,
      "grad_norm": 1.352597951889038,
      "learning_rate": 0.00019315810902029786,
      "loss": 3.1079,
      "step": 345
    },
    {
      "epoch": 0.20958083832335328,
      "grad_norm": 1.5166900157928467,
      "learning_rate": 0.00019277309676484858,
      "loss": 3.1142,
      "step": 350
    },
    {
      "epoch": 0.2125748502994012,
      "grad_norm": 1.1076050996780396,
      "learning_rate": 0.0001923779514662154,
      "loss": 3.0817,
      "step": 355
    },
    {
      "epoch": 0.2155688622754491,
      "grad_norm": 1.1077985763549805,
      "learning_rate": 0.00019197271628372482,
      "loss": 3.0804,
      "step": 360
    },
    {
      "epoch": 0.218562874251497,
      "grad_norm": 1.0608789920806885,
      "learning_rate": 0.00019155743547876023,
      "loss": 3.0823,
      "step": 365
    },
    {
      "epoch": 0.2215568862275449,
      "grad_norm": 1.2678635120391846,
      "learning_rate": 0.00019113215440992752,
      "loss": 3.0845,
      "step": 370
    },
    {
      "epoch": 0.2245508982035928,
      "grad_norm": 1.1210081577301025,
      "learning_rate": 0.0001906969195281007,
      "loss": 3.066,
      "step": 375
    },
    {
      "epoch": 0.2275449101796407,
      "grad_norm": 1.3343095779418945,
      "learning_rate": 0.00019025177837134858,
      "loss": 3.0512,
      "step": 380
    },
    {
      "epoch": 0.23053892215568864,
      "grad_norm": 1.1427243947982788,
      "learning_rate": 0.00018979677955974228,
      "loss": 3.073,
      "step": 385
    },
    {
      "epoch": 0.23353293413173654,
      "grad_norm": 1.2000013589859009,
      "learning_rate": 0.0001893319727900448,
      "loss": 3.0551,
      "step": 390
    },
    {
      "epoch": 0.23652694610778444,
      "grad_norm": 1.0942010879516602,
      "learning_rate": 0.0001888574088302831,
      "loss": 3.0414,
      "step": 395
    },
    {
      "epoch": 0.23952095808383234,
      "grad_norm": 0.9435099363327026,
      "learning_rate": 0.00018837313951420272,
      "loss": 3.0231,
      "step": 400
    },
    {
      "epoch": 0.24251497005988024,
      "grad_norm": 1.0394734144210815,
      "learning_rate": 0.00018787921773560657,
      "loss": 3.0175,
      "step": 405
    },
    {
      "epoch": 0.24550898203592814,
      "grad_norm": 0.9441319704055786,
      "learning_rate": 0.00018737569744257756,
      "loss": 3.0385,
      "step": 410
    },
    {
      "epoch": 0.24850299401197604,
      "grad_norm": 1.020687460899353,
      "learning_rate": 0.00018686263363158602,
      "loss": 3.0505,
      "step": 415
    },
    {
      "epoch": 0.25149700598802394,
      "grad_norm": 0.9998067617416382,
      "learning_rate": 0.0001863400823414831,
      "loss": 3.0358,
      "step": 420
    },
    {
      "epoch": 0.25449101796407186,
      "grad_norm": 1.0903741121292114,
      "learning_rate": 0.00018580810064737965,
      "loss": 3.0195,
      "step": 425
    },
    {
      "epoch": 0.25748502994011974,
      "grad_norm": 1.0015404224395752,
      "learning_rate": 0.00018526674665441257,
      "loss": 3.0362,
      "step": 430
    },
    {
      "epoch": 0.26047904191616766,
      "grad_norm": 0.9371241927146912,
      "learning_rate": 0.00018471607949139803,
      "loss": 3.0123,
      "step": 435
    },
    {
      "epoch": 0.2634730538922156,
      "grad_norm": 0.9184331893920898,
      "learning_rate": 0.00018415615930437337,
      "loss": 3.0075,
      "step": 440
    },
    {
      "epoch": 0.26646706586826346,
      "grad_norm": 1.0484634637832642,
      "learning_rate": 0.00018358704725002768,
      "loss": 3.0029,
      "step": 445
    },
    {
      "epoch": 0.2694610778443114,
      "grad_norm": 1.0423493385314941,
      "learning_rate": 0.000183008805489022,
      "loss": 2.9913,
      "step": 450
    },
    {
      "epoch": 0.27245508982035926,
      "grad_norm": 1.2088091373443604,
      "learning_rate": 0.00018242149717919993,
      "loss": 2.9834,
      "step": 455
    },
    {
      "epoch": 0.2754491017964072,
      "grad_norm": 0.8996514678001404,
      "learning_rate": 0.0001818251864686893,
      "loss": 2.972,
      "step": 460
    },
    {
      "epoch": 0.27844311377245506,
      "grad_norm": 1.014440894126892,
      "learning_rate": 0.00018121993848889552,
      "loss": 2.9687,
      "step": 465
    },
    {
      "epoch": 0.281437125748503,
      "grad_norm": 0.8995974063873291,
      "learning_rate": 0.00018060581934738784,
      "loss": 2.9317,
      "step": 470
    },
    {
      "epoch": 0.2844311377245509,
      "grad_norm": 1.0253899097442627,
      "learning_rate": 0.00017998289612067864,
      "loss": 2.9695,
      "step": 475
    },
    {
      "epoch": 0.2874251497005988,
      "grad_norm": 0.954319179058075,
      "learning_rate": 0.00017935123684689733,
      "loss": 2.9884,
      "step": 480
    },
    {
      "epoch": 0.2904191616766467,
      "grad_norm": 0.8713873028755188,
      "learning_rate": 0.00017871091051835874,
      "loss": 2.9689,
      "step": 485
    },
    {
      "epoch": 0.2934131736526946,
      "grad_norm": 0.9465665817260742,
      "learning_rate": 0.00017806198707402752,
      "loss": 2.9073,
      "step": 490
    },
    {
      "epoch": 0.2964071856287425,
      "grad_norm": 0.9693202972412109,
      "learning_rate": 0.00017740453739187922,
      "loss": 2.953,
      "step": 495
    },
    {
      "epoch": 0.2994011976047904,
      "grad_norm": 0.9308465719223022,
      "learning_rate": 0.0001767386332811587,
      "loss": 2.9439,
      "step": 500
    },
    {
      "epoch": 0.3023952095808383,
      "grad_norm": 0.8614184260368347,
      "learning_rate": 0.0001760643474745368,
      "loss": 2.9567,
      "step": 505
    },
    {
      "epoch": 0.30538922155688625,
      "grad_norm": 0.913870632648468,
      "learning_rate": 0.00017538175362016622,
      "loss": 2.9447,
      "step": 510
    },
    {
      "epoch": 0.3083832335329341,
      "grad_norm": 0.8058397769927979,
      "learning_rate": 0.00017469092627363738,
      "loss": 2.9677,
      "step": 515
    },
    {
      "epoch": 0.31137724550898205,
      "grad_norm": 0.7762190699577332,
      "learning_rate": 0.00017399194088983511,
      "loss": 2.9242,
      "step": 520
    },
    {
      "epoch": 0.3143712574850299,
      "grad_norm": 0.8444433808326721,
      "learning_rate": 0.0001732848738146973,
      "loss": 2.9612,
      "step": 525
    },
    {
      "epoch": 0.31736526946107785,
      "grad_norm": 0.7365703582763672,
      "learning_rate": 0.00017256980227687595,
      "loss": 2.9629,
      "step": 530
    },
    {
      "epoch": 0.3203592814371258,
      "grad_norm": 0.841096818447113,
      "learning_rate": 0.00017184680437930198,
      "loss": 2.9349,
      "step": 535
    },
    {
      "epoch": 0.32335329341317365,
      "grad_norm": 0.7778092622756958,
      "learning_rate": 0.00017111595909065466,
      "loss": 2.9147,
      "step": 540
    },
    {
      "epoch": 0.3263473053892216,
      "grad_norm": 0.8155597448348999,
      "learning_rate": 0.00017037734623673615,
      "loss": 2.9255,
      "step": 545
    },
    {
      "epoch": 0.32934131736526945,
      "grad_norm": 0.693092942237854,
      "learning_rate": 0.00016963104649175272,
      "loss": 2.902,
      "step": 550
    },
    {
      "epoch": 0.3323353293413174,
      "grad_norm": 0.8276283144950867,
      "learning_rate": 0.0001688771413695032,
      "loss": 2.9414,
      "step": 555
    },
    {
      "epoch": 0.33532934131736525,
      "grad_norm": 0.8656878471374512,
      "learning_rate": 0.00016811571321447566,
      "loss": 2.9105,
      "step": 560
    },
    {
      "epoch": 0.3383233532934132,
      "grad_norm": 0.713503897190094,
      "learning_rate": 0.00016734684519285344,
      "loss": 2.9138,
      "step": 565
    },
    {
      "epoch": 0.3413173652694611,
      "grad_norm": 0.7443877458572388,
      "learning_rate": 0.00016657062128343144,
      "loss": 2.8801,
      "step": 570
    },
    {
      "epoch": 0.344311377245509,
      "grad_norm": 0.8257678151130676,
      "learning_rate": 0.00016578712626844365,
      "loss": 2.8803,
      "step": 575
    },
    {
      "epoch": 0.3473053892215569,
      "grad_norm": 0.8291301131248474,
      "learning_rate": 0.00016499644572430278,
      "loss": 2.9076,
      "step": 580
    },
    {
      "epoch": 0.3502994011976048,
      "grad_norm": 0.738528311252594,
      "learning_rate": 0.0001641986660122534,
      "loss": 2.8978,
      "step": 585
    },
    {
      "epoch": 0.3532934131736527,
      "grad_norm": 0.7673987150192261,
      "learning_rate": 0.00016339387426893918,
      "loss": 2.9111,
      "step": 590
    },
    {
      "epoch": 0.3562874251497006,
      "grad_norm": 0.8196045756340027,
      "learning_rate": 0.0001625821583968855,
      "loss": 2.8467,
      "step": 595
    },
    {
      "epoch": 0.3592814371257485,
      "grad_norm": 0.7220514416694641,
      "learning_rate": 0.00016176360705489823,
      "loss": 2.8808,
      "step": 600
    },
    {
      "epoch": 0.36227544910179643,
      "grad_norm": 0.8424475193023682,
      "learning_rate": 0.00016093830964838035,
      "loss": 2.858,
      "step": 605
    },
    {
      "epoch": 0.3652694610778443,
      "grad_norm": 0.7498836517333984,
      "learning_rate": 0.00016010635631956652,
      "loss": 2.8987,
      "step": 610
    },
    {
      "epoch": 0.36826347305389223,
      "grad_norm": 0.7145126461982727,
      "learning_rate": 0.0001592678379376775,
      "loss": 2.8769,
      "step": 615
    },
    {
      "epoch": 0.3712574850299401,
      "grad_norm": 0.6803230047225952,
      "learning_rate": 0.0001584228460889949,
      "loss": 2.8812,
      "step": 620
    },
    {
      "epoch": 0.37425149700598803,
      "grad_norm": 0.6997132897377014,
      "learning_rate": 0.00015757147306685808,
      "loss": 2.8886,
      "step": 625
    },
    {
      "epoch": 0.3772455089820359,
      "grad_norm": 0.7119357585906982,
      "learning_rate": 0.00015671381186158312,
      "loss": 2.9011,
      "step": 630
    },
    {
      "epoch": 0.38023952095808383,
      "grad_norm": 0.8835166096687317,
      "learning_rate": 0.00015584995615030634,
      "loss": 2.8876,
      "step": 635
    },
    {
      "epoch": 0.38323353293413176,
      "grad_norm": 0.7940571904182434,
      "learning_rate": 0.0001549800002867524,
      "loss": 2.8739,
      "step": 640
    },
    {
      "epoch": 0.38622754491017963,
      "grad_norm": 0.7138169407844543,
      "learning_rate": 0.00015410403929092857,
      "loss": 2.869,
      "step": 645
    },
    {
      "epoch": 0.38922155688622756,
      "grad_norm": 0.711258053779602,
      "learning_rate": 0.00015322216883874643,
      "loss": 2.878,
      "step": 650
    },
    {
      "epoch": 0.39221556886227543,
      "grad_norm": 0.730831503868103,
      "learning_rate": 0.0001523344852515716,
      "loss": 2.8573,
      "step": 655
    },
    {
      "epoch": 0.39520958083832336,
      "grad_norm": 0.7164433598518372,
      "learning_rate": 0.00015144108548570322,
      "loss": 2.8375,
      "step": 660
    },
    {
      "epoch": 0.39820359281437123,
      "grad_norm": 0.7521623969078064,
      "learning_rate": 0.000150542067121784,
      "loss": 2.8714,
      "step": 665
    },
    {
      "epoch": 0.40119760479041916,
      "grad_norm": 0.6829126477241516,
      "learning_rate": 0.00014963752835414203,
      "loss": 2.8614,
      "step": 670
    },
    {
      "epoch": 0.4041916167664671,
      "grad_norm": 0.6588369011878967,
      "learning_rate": 0.00014872756798006576,
      "loss": 2.8498,
      "step": 675
    },
    {
      "epoch": 0.40718562874251496,
      "grad_norm": 0.670486569404602,
      "learning_rate": 0.00014781228538901267,
      "loss": 2.853,
      "step": 680
    },
    {
      "epoch": 0.4101796407185629,
      "grad_norm": 0.6251630783081055,
      "learning_rate": 0.00014689178055175394,
      "loss": 2.8709,
      "step": 685
    },
    {
      "epoch": 0.41317365269461076,
      "grad_norm": 0.6421228647232056,
      "learning_rate": 0.00014596615400945496,
      "loss": 2.8585,
      "step": 690
    },
    {
      "epoch": 0.4161676646706587,
      "grad_norm": 0.7128710746765137,
      "learning_rate": 0.0001450355068626939,
      "loss": 2.8333,
      "step": 695
    },
    {
      "epoch": 0.41916167664670656,
      "grad_norm": 0.7639265060424805,
      "learning_rate": 0.0001440999407604192,
      "loss": 2.8495,
      "step": 700
    },
    {
      "epoch": 0.4221556886227545,
      "grad_norm": 0.7937913537025452,
      "learning_rate": 0.00014315955788884698,
      "loss": 2.8667,
      "step": 705
    },
    {
      "epoch": 0.4251497005988024,
      "grad_norm": 0.7219135761260986,
      "learning_rate": 0.00014221446096029992,
      "loss": 2.8137,
      "step": 710
    },
    {
      "epoch": 0.4281437125748503,
      "grad_norm": 0.6893067359924316,
      "learning_rate": 0.00014126475320198843,
      "loss": 2.8147,
      "step": 715
    },
    {
      "epoch": 0.4311377245508982,
      "grad_norm": 0.7299759387969971,
      "learning_rate": 0.00014031053834473613,
      "loss": 2.8157,
      "step": 720
    },
    {
      "epoch": 0.4341317365269461,
      "grad_norm": 0.6203035116195679,
      "learning_rate": 0.00013935192061164956,
      "loss": 2.842,
      "step": 725
    },
    {
      "epoch": 0.437125748502994,
      "grad_norm": 0.67593914270401,
      "learning_rate": 0.0001383890047067348,
      "loss": 2.8281,
      "step": 730
    },
    {
      "epoch": 0.44011976047904194,
      "grad_norm": 0.629625678062439,
      "learning_rate": 0.0001374218958034612,
      "loss": 2.8321,
      "step": 735
    },
    {
      "epoch": 0.4431137724550898,
      "grad_norm": 0.7435698509216309,
      "learning_rate": 0.0001364506995332739,
      "loss": 2.8069,
      "step": 740
    },
    {
      "epoch": 0.44610778443113774,
      "grad_norm": 0.7014979124069214,
      "learning_rate": 0.00013547552197405632,
      "loss": 2.8271,
      "step": 745
    },
    {
      "epoch": 0.4491017964071856,
      "grad_norm": 0.6280860900878906,
      "learning_rate": 0.00013449646963854396,
      "loss": 2.7851,
      "step": 750
    },
    {
      "epoch": 0.45209580838323354,
      "grad_norm": 0.6429287195205688,
      "learning_rate": 0.00013351364946269072,
      "loss": 2.8342,
      "step": 755
    },
    {
      "epoch": 0.4550898203592814,
      "grad_norm": 0.6879150867462158,
      "learning_rate": 0.00013252716879398884,
      "loss": 2.8017,
      "step": 760
    },
    {
      "epoch": 0.45808383233532934,
      "grad_norm": 0.6519134044647217,
      "learning_rate": 0.00013153713537974394,
      "loss": 2.8272,
      "step": 765
    },
    {
      "epoch": 0.46107784431137727,
      "grad_norm": 0.7050421833992004,
      "learning_rate": 0.00013054365735530664,
      "loss": 2.8046,
      "step": 770
    },
    {
      "epoch": 0.46407185628742514,
      "grad_norm": 0.6754117012023926,
      "learning_rate": 0.00012954684323226136,
      "loss": 2.7996,
      "step": 775
    },
    {
      "epoch": 0.46706586826347307,
      "grad_norm": 0.6119791269302368,
      "learning_rate": 0.00012854680188657437,
      "loss": 2.825,
      "step": 780
    },
    {
      "epoch": 0.47005988023952094,
      "grad_norm": 0.7737504839897156,
      "learning_rate": 0.00012754364254670192,
      "loss": 2.8229,
      "step": 785
    },
    {
      "epoch": 0.47305389221556887,
      "grad_norm": 0.6356379985809326,
      "learning_rate": 0.00012653747478165987,
      "loss": 2.7901,
      "step": 790
    },
    {
      "epoch": 0.47604790419161674,
      "grad_norm": 0.679063081741333,
      "learning_rate": 0.0001255284084890562,
      "loss": 2.7847,
      "step": 795
    },
    {
      "epoch": 0.47904191616766467,
      "grad_norm": 0.6137480139732361,
      "learning_rate": 0.0001245165538830873,
      "loss": 2.7921,
      "step": 800
    },
    {
      "epoch": 0.4820359281437126,
      "grad_norm": 0.6218224167823792,
      "learning_rate": 0.00012350202148250037,
      "loss": 2.8368,
      "step": 805
    },
    {
      "epoch": 0.48502994011976047,
      "grad_norm": 0.6735543012619019,
      "learning_rate": 0.0001224849220985218,
      "loss": 2.7808,
      "step": 810
    },
    {
      "epoch": 0.4880239520958084,
      "grad_norm": 0.5928522944450378,
      "learning_rate": 0.00012146536682275387,
      "loss": 2.7922,
      "step": 815
    },
    {
      "epoch": 0.49101796407185627,
      "grad_norm": 0.640355110168457,
      "learning_rate": 0.00012044346701504128,
      "loss": 2.8388,
      "step": 820
    },
    {
      "epoch": 0.4940119760479042,
      "grad_norm": 0.5893163084983826,
      "learning_rate": 0.00011941933429130758,
      "loss": 2.8005,
      "step": 825
    },
    {
      "epoch": 0.49700598802395207,
      "grad_norm": 0.5506909489631653,
      "learning_rate": 0.0001183930805113643,
      "loss": 2.8018,
      "step": 830
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.6417153477668762,
      "learning_rate": 0.00011736481776669306,
      "loss": 2.7829,
      "step": 835
    },
    {
      "epoch": 0.5029940119760479,
      "grad_norm": 0.6764044761657715,
      "learning_rate": 0.00011633465836820243,
      "loss": 2.7853,
      "step": 840
    },
    {
      "epoch": 0.5059880239520959,
      "grad_norm": 0.6103654503822327,
      "learning_rate": 0.00011530271483396115,
      "loss": 2.7922,
      "step": 845
    },
    {
      "epoch": 0.5089820359281437,
      "grad_norm": 0.6200462579727173,
      "learning_rate": 0.00011426909987690819,
      "loss": 2.798,
      "step": 850
    },
    {
      "epoch": 0.5119760479041916,
      "grad_norm": 0.556036114692688,
      "learning_rate": 0.00011323392639254193,
      "loss": 2.823,
      "step": 855
    },
    {
      "epoch": 0.5149700598802395,
      "grad_norm": 0.5561334490776062,
      "learning_rate": 0.00011219730744658921,
      "loss": 2.7946,
      "step": 860
    },
    {
      "epoch": 0.5179640718562875,
      "grad_norm": 0.6291623711585999,
      "learning_rate": 0.00011115935626265594,
      "loss": 2.769,
      "step": 865
    },
    {
      "epoch": 0.5209580838323353,
      "grad_norm": 0.622279703617096,
      "learning_rate": 0.00011012018620986028,
      "loss": 2.7905,
      "step": 870
    },
    {
      "epoch": 0.5239520958083832,
      "grad_norm": 0.5651618242263794,
      "learning_rate": 0.00010907991079045006,
      "loss": 2.7611,
      "step": 875
    },
    {
      "epoch": 0.5269461077844312,
      "grad_norm": 0.5883045792579651,
      "learning_rate": 0.00010803864362740562,
      "loss": 2.7853,
      "step": 880
    },
    {
      "epoch": 0.5299401197604791,
      "grad_norm": 0.5984832048416138,
      "learning_rate": 0.00010699649845202934,
      "loss": 2.7686,
      "step": 885
    },
    {
      "epoch": 0.5329341317365269,
      "grad_norm": 0.631034791469574,
      "learning_rate": 0.00010595358909152378,
      "loss": 2.7649,
      "step": 890
    },
    {
      "epoch": 0.5359281437125748,
      "grad_norm": 0.5855838656425476,
      "learning_rate": 0.00010491002945655861,
      "loss": 2.7972,
      "step": 895
    },
    {
      "epoch": 0.5389221556886228,
      "grad_norm": 0.644307017326355,
      "learning_rate": 0.00010386593352882909,
      "loss": 2.7894,
      "step": 900
    },
    {
      "epoch": 0.5419161676646707,
      "grad_norm": 0.6458478569984436,
      "learning_rate": 0.0001028214153486066,
      "loss": 2.7875,
      "step": 905
    },
    {
      "epoch": 0.5449101796407185,
      "grad_norm": 0.6464332342147827,
      "learning_rate": 0.00010177658900228249,
      "loss": 2.7967,
      "step": 910
    },
    {
      "epoch": 0.5479041916167665,
      "grad_norm": 0.564663827419281,
      "learning_rate": 0.0001007315686099072,
      "loss": 2.8149,
      "step": 915
    },
    {
      "epoch": 0.5508982035928144,
      "grad_norm": 0.669620931148529,
      "learning_rate": 9.96864683127257e-05,
      "loss": 2.773,
      "step": 920
    },
    {
      "epoch": 0.5538922155688623,
      "grad_norm": 0.6294126510620117,
      "learning_rate": 9.864140226071053e-05,
      "loss": 2.7909,
      "step": 925
    },
    {
      "epoch": 0.5568862275449101,
      "grad_norm": 0.5996381044387817,
      "learning_rate": 9.759648460009376e-05,
      "loss": 2.7736,
      "step": 930
    },
    {
      "epoch": 0.5598802395209581,
      "grad_norm": 0.6062589287757874,
      "learning_rate": 9.655182946089956e-05,
      "loss": 2.7693,
      "step": 935
    },
    {
      "epoch": 0.562874251497006,
      "grad_norm": 0.6130212545394897,
      "learning_rate": 9.550755094447848e-05,
      "loss": 2.7422,
      "step": 940
    },
    {
      "epoch": 0.5658682634730539,
      "grad_norm": 0.6118180155754089,
      "learning_rate": 9.446376311104494e-05,
      "loss": 2.7847,
      "step": 945
    },
    {
      "epoch": 0.5688622754491018,
      "grad_norm": 0.6395862698554993,
      "learning_rate": 9.342057996721894e-05,
      "loss": 2.7557,
      "step": 950
    },
    {
      "epoch": 0.5718562874251497,
      "grad_norm": 0.5906466841697693,
      "learning_rate": 9.237811545357392e-05,
      "loss": 2.7821,
      "step": 955
    },
    {
      "epoch": 0.5748502994011976,
      "grad_norm": 0.6405777335166931,
      "learning_rate": 9.133648343219168e-05,
      "loss": 2.7561,
      "step": 960
    },
    {
      "epoch": 0.5778443113772455,
      "grad_norm": 0.537714421749115,
      "learning_rate": 9.029579767422592e-05,
      "loss": 2.7624,
      "step": 965
    },
    {
      "epoch": 0.5808383233532934,
      "grad_norm": 0.5861092805862427,
      "learning_rate": 8.925617184747584e-05,
      "loss": 2.7885,
      "step": 970
    },
    {
      "epoch": 0.5838323353293413,
      "grad_norm": 0.5366087555885315,
      "learning_rate": 8.821771950397066e-05,
      "loss": 2.7717,
      "step": 975
    },
    {
      "epoch": 0.5868263473053892,
      "grad_norm": 0.5149749517440796,
      "learning_rate": 8.718055406756714e-05,
      "loss": 2.7722,
      "step": 980
    },
    {
      "epoch": 0.5898203592814372,
      "grad_norm": 0.6070750951766968,
      "learning_rate": 8.614478882156103e-05,
      "loss": 2.7747,
      "step": 985
    },
    {
      "epoch": 0.592814371257485,
      "grad_norm": 0.5566183924674988,
      "learning_rate": 8.51105368963137e-05,
      "loss": 2.7363,
      "step": 990
    },
    {
      "epoch": 0.5958083832335329,
      "grad_norm": 0.5821384191513062,
      "learning_rate": 8.407791125689578e-05,
      "loss": 2.7286,
      "step": 995
    },
    {
      "epoch": 0.5988023952095808,
      "grad_norm": 0.5972596406936646,
      "learning_rate": 8.30470246907484e-05,
      "loss": 2.8008,
      "step": 1000
    },
    {
      "epoch": 0.6017964071856288,
      "grad_norm": 0.5653948187828064,
      "learning_rate": 8.201798979536437e-05,
      "loss": 2.7381,
      "step": 1005
    },
    {
      "epoch": 0.6047904191616766,
      "grad_norm": 0.5987522602081299,
      "learning_rate": 8.099091896598964e-05,
      "loss": 2.7171,
      "step": 1010
    },
    {
      "epoch": 0.6077844311377245,
      "grad_norm": 0.6162002086639404,
      "learning_rate": 7.996592438334728e-05,
      "loss": 2.7595,
      "step": 1015
    },
    {
      "epoch": 0.6107784431137725,
      "grad_norm": 0.5301627516746521,
      "learning_rate": 7.894311800138432e-05,
      "loss": 2.7199,
      "step": 1020
    },
    {
      "epoch": 0.6137724550898204,
      "grad_norm": 0.5902780890464783,
      "learning_rate": 7.792261153504402e-05,
      "loss": 2.7465,
      "step": 1025
    },
    {
      "epoch": 0.6167664670658682,
      "grad_norm": 0.5662907958030701,
      "learning_rate": 7.690451644806372e-05,
      "loss": 2.7463,
      "step": 1030
    },
    {
      "epoch": 0.6197604790419161,
      "grad_norm": 0.5153141617774963,
      "learning_rate": 7.588894394080045e-05,
      "loss": 2.7497,
      "step": 1035
    },
    {
      "epoch": 0.6227544910179641,
      "grad_norm": 0.5608941316604614,
      "learning_rate": 7.487600493808513e-05,
      "loss": 2.7587,
      "step": 1040
    },
    {
      "epoch": 0.625748502994012,
      "grad_norm": 0.5900591015815735,
      "learning_rate": 7.386581007710693e-05,
      "loss": 2.721,
      "step": 1045
    },
    {
      "epoch": 0.6287425149700598,
      "grad_norm": 0.5923062562942505,
      "learning_rate": 7.285846969532907e-05,
      "loss": 2.7245,
      "step": 1050
    },
    {
      "epoch": 0.6317365269461078,
      "grad_norm": 0.5622250437736511,
      "learning_rate": 7.185409381843727e-05,
      "loss": 2.7514,
      "step": 1055
    },
    {
      "epoch": 0.6347305389221557,
      "grad_norm": 0.5878713130950928,
      "learning_rate": 7.085279214832233e-05,
      "loss": 2.7361,
      "step": 1060
    },
    {
      "epoch": 0.6377245508982036,
      "grad_norm": 0.5391808748245239,
      "learning_rate": 6.985467405109815e-05,
      "loss": 2.7708,
      "step": 1065
    },
    {
      "epoch": 0.6407185628742516,
      "grad_norm": 0.5266147255897522,
      "learning_rate": 6.885984854515623e-05,
      "loss": 2.739,
      "step": 1070
    },
    {
      "epoch": 0.6437125748502994,
      "grad_norm": 0.5293008089065552,
      "learning_rate": 6.786842428925821e-05,
      "loss": 2.7338,
      "step": 1075
    },
    {
      "epoch": 0.6467065868263473,
      "grad_norm": 0.5374533534049988,
      "learning_rate": 6.688050957066787e-05,
      "loss": 2.7521,
      "step": 1080
    },
    {
      "epoch": 0.6497005988023952,
      "grad_norm": 0.519822895526886,
      "learning_rate": 6.58962122933234e-05,
      "loss": 2.7509,
      "step": 1085
    },
    {
      "epoch": 0.6526946107784432,
      "grad_norm": 0.4745498597621918,
      "learning_rate": 6.491563996605198e-05,
      "loss": 2.6944,
      "step": 1090
    },
    {
      "epoch": 0.655688622754491,
      "grad_norm": 0.5428311228752136,
      "learning_rate": 6.393889969082691e-05,
      "loss": 2.746,
      "step": 1095
    },
    {
      "epoch": 0.6586826347305389,
      "grad_norm": 0.6130102276802063,
      "learning_rate": 6.29660981510697e-05,
      "loss": 2.7142,
      "step": 1100
    },
    {
      "epoch": 0.6616766467065869,
      "grad_norm": 0.565981388092041,
      "learning_rate": 6.199734159999769e-05,
      "loss": 2.7377,
      "step": 1105
    },
    {
      "epoch": 0.6646706586826348,
      "grad_norm": 0.5873706340789795,
      "learning_rate": 6.103273584901856e-05,
      "loss": 2.7707,
      "step": 1110
    },
    {
      "epoch": 0.6676646706586826,
      "grad_norm": 0.6199204325675964,
      "learning_rate": 6.007238625617333e-05,
      "loss": 2.7288,
      "step": 1115
    },
    {
      "epoch": 0.6706586826347305,
      "grad_norm": 0.5575565695762634,
      "learning_rate": 5.911639771462858e-05,
      "loss": 2.7454,
      "step": 1120
    },
    {
      "epoch": 0.6736526946107785,
      "grad_norm": 0.5518234372138977,
      "learning_rate": 5.8164874641219735e-05,
      "loss": 2.7345,
      "step": 1125
    },
    {
      "epoch": 0.6766467065868264,
      "grad_norm": 0.5617517828941345,
      "learning_rate": 5.721792096504611e-05,
      "loss": 2.756,
      "step": 1130
    },
    {
      "epoch": 0.6796407185628742,
      "grad_norm": 0.5457236170768738,
      "learning_rate": 5.627564011611961e-05,
      "loss": 2.7492,
      "step": 1135
    },
    {
      "epoch": 0.6826347305389222,
      "grad_norm": 0.5588122606277466,
      "learning_rate": 5.5338135014067395e-05,
      "loss": 2.7385,
      "step": 1140
    },
    {
      "epoch": 0.6856287425149701,
      "grad_norm": 0.5922512412071228,
      "learning_rate": 5.440550805689075e-05,
      "loss": 2.7128,
      "step": 1145
    },
    {
      "epoch": 0.688622754491018,
      "grad_norm": 0.5327529311180115,
      "learning_rate": 5.3477861109780835e-05,
      "loss": 2.7507,
      "step": 1150
    },
    {
      "epoch": 0.6916167664670658,
      "grad_norm": 0.5234887003898621,
      "learning_rate": 5.255529549399234e-05,
      "loss": 2.77,
      "step": 1155
    },
    {
      "epoch": 0.6946107784431138,
      "grad_norm": 0.5305109024047852,
      "learning_rate": 5.163791197577714e-05,
      "loss": 2.7765,
      "step": 1160
    },
    {
      "epoch": 0.6976047904191617,
      "grad_norm": 0.5282033085823059,
      "learning_rate": 5.0725810755377825e-05,
      "loss": 2.7518,
      "step": 1165
    },
    {
      "epoch": 0.7005988023952096,
      "grad_norm": 0.5210415124893188,
      "learning_rate": 4.9819091456083644e-05,
      "loss": 2.7493,
      "step": 1170
    },
    {
      "epoch": 0.7035928143712575,
      "grad_norm": 0.560530424118042,
      "learning_rate": 4.891785311334923e-05,
      "loss": 2.7429,
      "step": 1175
    },
    {
      "epoch": 0.7065868263473054,
      "grad_norm": 0.5484561324119568,
      "learning_rate": 4.8022194163977494e-05,
      "loss": 2.7131,
      "step": 1180
    },
    {
      "epoch": 0.7095808383233533,
      "grad_norm": 0.5590568780899048,
      "learning_rate": 4.713221243536816e-05,
      "loss": 2.715,
      "step": 1185
    },
    {
      "epoch": 0.7125748502994012,
      "grad_norm": 0.5286269783973694,
      "learning_rate": 4.6248005134832394e-05,
      "loss": 2.7006,
      "step": 1190
    },
    {
      "epoch": 0.7155688622754491,
      "grad_norm": 0.5053289532661438,
      "learning_rate": 4.5369668838975597e-05,
      "loss": 2.7375,
      "step": 1195
    },
    {
      "epoch": 0.718562874251497,
      "grad_norm": 0.5476385354995728,
      "learning_rate": 4.449729948314894e-05,
      "loss": 2.7402,
      "step": 1200
    },
    {
      "epoch": 0.7215568862275449,
      "grad_norm": 0.5178956985473633,
      "learning_rate": 4.363099235097087e-05,
      "loss": 2.7,
      "step": 1205
    },
    {
      "epoch": 0.7245508982035929,
      "grad_norm": 0.5388510227203369,
      "learning_rate": 4.277084206391989e-05,
      "loss": 2.7014,
      "step": 1210
    },
    {
      "epoch": 0.7275449101796407,
      "grad_norm": 0.510368824005127,
      "learning_rate": 4.191694257099962e-05,
      "loss": 2.7267,
      "step": 1215
    },
    {
      "epoch": 0.7305389221556886,
      "grad_norm": 0.563225269317627,
      "learning_rate": 4.10693871384773e-05,
      "loss": 2.7277,
      "step": 1220
    },
    {
      "epoch": 0.7335329341317365,
      "grad_norm": 0.5411152243614197,
      "learning_rate": 4.022826833969692e-05,
      "loss": 2.7442,
      "step": 1225
    },
    {
      "epoch": 0.7365269461077845,
      "grad_norm": 0.5123066902160645,
      "learning_rate": 3.93936780449679e-05,
      "loss": 2.7042,
      "step": 1230
    },
    {
      "epoch": 0.7395209580838323,
      "grad_norm": 0.5124099254608154,
      "learning_rate": 3.856570741153087e-05,
      "loss": 2.7288,
      "step": 1235
    },
    {
      "epoch": 0.7425149700598802,
      "grad_norm": 0.5201666355133057,
      "learning_rate": 3.774444687360082e-05,
      "loss": 2.7266,
      "step": 1240
    },
    {
      "epoch": 0.7455089820359282,
      "grad_norm": 0.5414671301841736,
      "learning_rate": 3.692998613248977e-05,
      "loss": 2.7237,
      "step": 1245
    },
    {
      "epoch": 0.7485029940119761,
      "grad_norm": 0.5279539823532104,
      "learning_rate": 3.6122414146809014e-05,
      "loss": 2.7114,
      "step": 1250
    },
    {
      "epoch": 0.7514970059880239,
      "grad_norm": 0.5688903331756592,
      "learning_rate": 3.532181912275301e-05,
      "loss": 2.7434,
      "step": 1255
    },
    {
      "epoch": 0.7544910179640718,
      "grad_norm": 0.5434580445289612,
      "learning_rate": 3.4528288504464844e-05,
      "loss": 2.7369,
      "step": 1260
    },
    {
      "epoch": 0.7574850299401198,
      "grad_norm": 0.5412757992744446,
      "learning_rate": 3.3741908964485414e-05,
      "loss": 2.7672,
      "step": 1265
    },
    {
      "epoch": 0.7604790419161677,
      "grad_norm": 0.5109795331954956,
      "learning_rate": 3.296276639428665e-05,
      "loss": 2.6753,
      "step": 1270
    },
    {
      "epoch": 0.7634730538922155,
      "grad_norm": 0.5179950594902039,
      "learning_rate": 3.21909458948901e-05,
      "loss": 2.7301,
      "step": 1275
    },
    {
      "epoch": 0.7664670658682635,
      "grad_norm": 0.5099706649780273,
      "learning_rate": 3.1426531767572e-05,
      "loss": 2.7018,
      "step": 1280
    },
    {
      "epoch": 0.7694610778443114,
      "grad_norm": 0.5369354486465454,
      "learning_rate": 3.0669607504655326e-05,
      "loss": 2.7473,
      "step": 1285
    },
    {
      "epoch": 0.7724550898203593,
      "grad_norm": 0.45971599221229553,
      "learning_rate": 2.9920255780390617e-05,
      "loss": 2.7035,
      "step": 1290
    },
    {
      "epoch": 0.7754491017964071,
      "grad_norm": 0.49725160002708435,
      "learning_rate": 2.917855844192584e-05,
      "loss": 2.7258,
      "step": 1295
    },
    {
      "epoch": 0.7784431137724551,
      "grad_norm": 0.48002126812934875,
      "learning_rate": 2.8444596500366825e-05,
      "loss": 2.6843,
      "step": 1300
    },
    {
      "epoch": 0.781437125748503,
      "grad_norm": 0.47459498047828674,
      "learning_rate": 2.7718450121928918e-05,
      "loss": 2.7133,
      "step": 1305
    },
    {
      "epoch": 0.7844311377245509,
      "grad_norm": 0.5074918270111084,
      "learning_rate": 2.7000198619180794e-05,
      "loss": 2.7191,
      "step": 1310
    },
    {
      "epoch": 0.7874251497005988,
      "grad_norm": 0.48583555221557617,
      "learning_rate": 2.6289920442381722e-05,
      "loss": 2.7248,
      "step": 1315
    },
    {
      "epoch": 0.7904191616766467,
      "grad_norm": 0.5015809535980225,
      "learning_rate": 2.5587693170912875e-05,
      "loss": 2.7373,
      "step": 1320
    },
    {
      "epoch": 0.7934131736526946,
      "grad_norm": 0.49757909774780273,
      "learning_rate": 2.4893593504803826e-05,
      "loss": 2.7156,
      "step": 1325
    },
    {
      "epoch": 0.7964071856287425,
      "grad_norm": 0.48436880111694336,
      "learning_rate": 2.4207697256355145e-05,
      "loss": 2.7378,
      "step": 1330
    },
    {
      "epoch": 0.7994011976047904,
      "grad_norm": 0.5281265377998352,
      "learning_rate": 2.353007934185768e-05,
      "loss": 2.7332,
      "step": 1335
    },
    {
      "epoch": 0.8023952095808383,
      "grad_norm": 0.5300716161727905,
      "learning_rate": 2.2860813773410106e-05,
      "loss": 2.663,
      "step": 1340
    },
    {
      "epoch": 0.8053892215568862,
      "grad_norm": 0.429849237203598,
      "learning_rate": 2.2199973650834906e-05,
      "loss": 2.6993,
      "step": 1345
    },
    {
      "epoch": 0.8083832335329342,
      "grad_norm": 0.5136377811431885,
      "learning_rate": 2.154763115369419e-05,
      "loss": 2.7297,
      "step": 1350
    },
    {
      "epoch": 0.811377245508982,
      "grad_norm": 0.523758053779602,
      "learning_rate": 2.0903857533405958e-05,
      "loss": 2.7227,
      "step": 1355
    },
    {
      "epoch": 0.8143712574850299,
      "grad_norm": 0.5091609954833984,
      "learning_rate": 2.026872310546165e-05,
      "loss": 2.7174,
      "step": 1360
    },
    {
      "epoch": 0.8173652694610778,
      "grad_norm": 0.5340681672096252,
      "learning_rate": 1.9642297241746142e-05,
      "loss": 2.706,
      "step": 1365
    },
    {
      "epoch": 0.8203592814371258,
      "grad_norm": 0.4736381471157074,
      "learning_rate": 1.902464836296054e-05,
      "loss": 2.7244,
      "step": 1370
    },
    {
      "epoch": 0.8233532934131736,
      "grad_norm": 0.5413190126419067,
      "learning_rate": 1.841584393114919e-05,
      "loss": 2.682,
      "step": 1375
    },
    {
      "epoch": 0.8263473053892215,
      "grad_norm": 0.45962581038475037,
      "learning_rate": 1.7815950442330963e-05,
      "loss": 2.6946,
      "step": 1380
    },
    {
      "epoch": 0.8293413173652695,
      "grad_norm": 0.5146113634109497,
      "learning_rate": 1.7225033419236503e-05,
      "loss": 2.6919,
      "step": 1385
    },
    {
      "epoch": 0.8323353293413174,
      "grad_norm": 0.48548585176467896,
      "learning_rate": 1.6643157404151467e-05,
      "loss": 2.7278,
      "step": 1390
    },
    {
      "epoch": 0.8353293413173652,
      "grad_norm": 0.46142202615737915,
      "learning_rate": 1.6070385951866952e-05,
      "loss": 2.7241,
      "step": 1395
    },
    {
      "epoch": 0.8383233532934131,
      "grad_norm": 0.5457370281219482,
      "learning_rate": 1.5506781622737942e-05,
      "loss": 2.7047,
      "step": 1400
    },
    {
      "epoch": 0.8413173652694611,
      "grad_norm": 0.48171794414520264,
      "learning_rate": 1.4952405975850026e-05,
      "loss": 2.6901,
      "step": 1405
    },
    {
      "epoch": 0.844311377245509,
      "grad_norm": 0.5125330090522766,
      "learning_rate": 1.4407319562295762e-05,
      "loss": 2.7052,
      "step": 1410
    },
    {
      "epoch": 0.8473053892215568,
      "grad_norm": 0.4971305727958679,
      "learning_rate": 1.387158191856105e-05,
      "loss": 2.7405,
      "step": 1415
    },
    {
      "epoch": 0.8502994011976048,
      "grad_norm": 0.539444625377655,
      "learning_rate": 1.3345251560022288e-05,
      "loss": 2.6869,
      "step": 1420
    },
    {
      "epoch": 0.8532934131736527,
      "grad_norm": 0.5079681873321533,
      "learning_rate": 1.2828385974555202e-05,
      "loss": 2.7132,
      "step": 1425
    },
    {
      "epoch": 0.8562874251497006,
      "grad_norm": 0.5335781574249268,
      "learning_rate": 1.2321041616255614e-05,
      "loss": 2.7378,
      "step": 1430
    },
    {
      "epoch": 0.8592814371257484,
      "grad_norm": 0.4930953085422516,
      "learning_rate": 1.1823273899273435e-05,
      "loss": 2.678,
      "step": 1435
    },
    {
      "epoch": 0.8622754491017964,
      "grad_norm": 0.4556470513343811,
      "learning_rate": 1.1335137191760093e-05,
      "loss": 2.6918,
      "step": 1440
    },
    {
      "epoch": 0.8652694610778443,
      "grad_norm": 0.49691587686538696,
      "learning_rate": 1.0856684809930151e-05,
      "loss": 2.6948,
      "step": 1445
    },
    {
      "epoch": 0.8682634730538922,
      "grad_norm": 0.48159509897232056,
      "learning_rate": 1.0387969012238064e-05,
      "loss": 2.694,
      "step": 1450
    },
    {
      "epoch": 0.8712574850299402,
      "grad_norm": 0.4907771050930023,
      "learning_rate": 9.929040993670114e-06,
      "loss": 2.724,
      "step": 1455
    },
    {
      "epoch": 0.874251497005988,
      "grad_norm": 0.4428238570690155,
      "learning_rate": 9.47995088015281e-06,
      "loss": 2.6853,
      "step": 1460
    },
    {
      "epoch": 0.8772455089820359,
      "grad_norm": 0.4720156192779541,
      "learning_rate": 9.040747723077902e-06,
      "loss": 2.7465,
      "step": 1465
    },
    {
      "epoch": 0.8802395209580839,
      "grad_norm": 0.4877772629261017,
      "learning_rate": 8.61147949394483e-06,
      "loss": 2.7284,
      "step": 1470
    },
    {
      "epoch": 0.8832335329341318,
      "grad_norm": 0.48317137360572815,
      "learning_rate": 8.192193079121002e-06,
      "loss": 2.7261,
      "step": 1475
    },
    {
      "epoch": 0.8862275449101796,
      "grad_norm": 0.48045286536216736,
      "learning_rate": 7.782934274720777e-06,
      "loss": 2.6818,
      "step": 1480
    },
    {
      "epoch": 0.8892215568862275,
      "grad_norm": 0.47768479585647583,
      "learning_rate": 7.3837477816033896e-06,
      "loss": 2.692,
      "step": 1485
    },
    {
      "epoch": 0.8922155688622755,
      "grad_norm": 0.5294510722160339,
      "learning_rate": 6.994677200490507e-06,
      "loss": 2.6831,
      "step": 1490
    },
    {
      "epoch": 0.8952095808383234,
      "grad_norm": 0.48493650555610657,
      "learning_rate": 6.615765027204102e-06,
      "loss": 2.6695,
      "step": 1495
    },
    {
      "epoch": 0.8982035928143712,
      "grad_norm": 0.5145286917686462,
      "learning_rate": 6.247052648024765e-06,
      "loss": 2.6966,
      "step": 1500
    },
    {
      "epoch": 0.9011976047904192,
      "grad_norm": 0.4654744565486908,
      "learning_rate": 5.888580335171368e-06,
      "loss": 2.7247,
      "step": 1505
    },
    {
      "epoch": 0.9041916167664671,
      "grad_norm": 0.4812209904193878,
      "learning_rate": 5.540387242402434e-06,
      "loss": 2.7261,
      "step": 1510
    },
    {
      "epoch": 0.907185628742515,
      "grad_norm": 0.4804396331310272,
      "learning_rate": 5.20251140073953e-06,
      "loss": 2.692,
      "step": 1515
    },
    {
      "epoch": 0.9101796407185628,
      "grad_norm": 0.4617965519428253,
      "learning_rate": 4.874989714313449e-06,
      "loss": 2.7015,
      "step": 1520
    },
    {
      "epoch": 0.9131736526946108,
      "grad_norm": 0.4632161259651184,
      "learning_rate": 4.5578579563333e-06,
      "loss": 2.6971,
      "step": 1525
    },
    {
      "epoch": 0.9161676646706587,
      "grad_norm": 0.530784010887146,
      "learning_rate": 4.251150765179291e-06,
      "loss": 2.7031,
      "step": 1530
    },
    {
      "epoch": 0.9191616766467066,
      "grad_norm": 0.49502164125442505,
      "learning_rate": 3.954901640619368e-06,
      "loss": 2.6704,
      "step": 1535
    },
    {
      "epoch": 0.9221556886227545,
      "grad_norm": 0.4807766079902649,
      "learning_rate": 3.6691429401502053e-06,
      "loss": 2.6889,
      "step": 1540
    },
    {
      "epoch": 0.9251497005988024,
      "grad_norm": 0.49187591671943665,
      "learning_rate": 3.3939058754630882e-06,
      "loss": 2.7056,
      "step": 1545
    },
    {
      "epoch": 0.9281437125748503,
      "grad_norm": 0.5204874277114868,
      "learning_rate": 3.1292205090347248e-06,
      "loss": 2.7579,
      "step": 1550
    },
    {
      "epoch": 0.9311377245508982,
      "grad_norm": 0.5218049883842468,
      "learning_rate": 2.875115750843771e-06,
      "loss": 2.722,
      "step": 1555
    },
    {
      "epoch": 0.9341317365269461,
      "grad_norm": 0.4557659327983856,
      "learning_rate": 2.6316193552131884e-06,
      "loss": 2.7076,
      "step": 1560
    },
    {
      "epoch": 0.937125748502994,
      "grad_norm": 0.4934738874435425,
      "learning_rate": 2.398757917778727e-06,
      "loss": 2.7231,
      "step": 1565
    },
    {
      "epoch": 0.9401197604790419,
      "grad_norm": 0.48129838705062866,
      "learning_rate": 2.176556872584168e-06,
      "loss": 2.7231,
      "step": 1570
    },
    {
      "epoch": 0.9431137724550899,
      "grad_norm": 0.4986419975757599,
      "learning_rate": 1.965040489303194e-06,
      "loss": 2.7142,
      "step": 1575
    },
    {
      "epoch": 0.9461077844311377,
      "grad_norm": 0.48590248823165894,
      "learning_rate": 1.7642318705886286e-06,
      "loss": 2.7517,
      "step": 1580
    },
    {
      "epoch": 0.9491017964071856,
      "grad_norm": 0.4665396809577942,
      "learning_rate": 1.574152949549057e-06,
      "loss": 2.7028,
      "step": 1585
    },
    {
      "epoch": 0.9520958083832335,
      "grad_norm": 0.4755544364452362,
      "learning_rate": 1.3948244873532078e-06,
      "loss": 2.6893,
      "step": 1590
    },
    {
      "epoch": 0.9550898203592815,
      "grad_norm": 0.44289329648017883,
      "learning_rate": 1.226266070962323e-06,
      "loss": 2.7409,
      "step": 1595
    },
    {
      "epoch": 0.9580838323353293,
      "grad_norm": 0.5039628744125366,
      "learning_rate": 1.0684961109908353e-06,
      "loss": 2.6975,
      "step": 1600
    },
    {
      "epoch": 0.9610778443113772,
      "grad_norm": 0.47264382243156433,
      "learning_rate": 9.21531839695411e-07,
      "loss": 2.6629,
      "step": 1605
    },
    {
      "epoch": 0.9640718562874252,
      "grad_norm": 0.4993440806865692,
      "learning_rate": 7.853893090928654e-07,
      "loss": 2.7854,
      "step": 1610
    },
    {
      "epoch": 0.9670658682634731,
      "grad_norm": 0.4866315722465515,
      "learning_rate": 6.600833892068336e-07,
      "loss": 2.7192,
      "step": 1615
    },
    {
      "epoch": 0.9700598802395209,
      "grad_norm": 0.42745715379714966,
      "learning_rate": 5.456277664436127e-07,
      "loss": 2.7148,
      "step": 1620
    },
    {
      "epoch": 0.9730538922155688,
      "grad_norm": 0.49950963258743286,
      "learning_rate": 4.4203494209733576e-07,
      "loss": 2.743,
      "step": 1625
    },
    {
      "epoch": 0.9760479041916168,
      "grad_norm": 0.45195409655570984,
      "learning_rate": 3.4931623098445334e-07,
      "loss": 2.7275,
      "step": 1630
    },
    {
      "epoch": 0.9790419161676647,
      "grad_norm": 0.47945547103881836,
      "learning_rate": 2.674817602079327e-07,
      "loss": 2.725,
      "step": 1635
    },
    {
      "epoch": 0.9820359281437125,
      "grad_norm": 0.48270925879478455,
      "learning_rate": 1.965404680511207e-07,
      "loss": 2.6793,
      "step": 1640
    },
    {
      "epoch": 0.9850299401197605,
      "grad_norm": 0.4643472731113434,
      "learning_rate": 1.3650010300150228e-07,
      "loss": 2.7078,
      "step": 1645
    },
    {
      "epoch": 0.9880239520958084,
      "grad_norm": 0.4867922365665436,
      "learning_rate": 8.736722290429988e-08,
      "loss": 2.7148,
      "step": 1650
    },
    {
      "epoch": 0.9910179640718563,
      "grad_norm": 0.5043019652366638,
      "learning_rate": 4.9147194246290664e-08,
      "loss": 2.7255,
      "step": 1655
    },
    {
      "epoch": 0.9940119760479041,
      "grad_norm": 0.4517767131328583,
      "learning_rate": 2.1844191569597716e-08,
      "loss": 2.6949,
      "step": 1660
    },
    {
      "epoch": 0.9970059880239521,
      "grad_norm": 0.4739466905593872,
      "learning_rate": 5.461197015765862e-09,
      "loss": 2.7152,
      "step": 1665
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.46331822872161865,
      "learning_rate": 0.0,
      "loss": 2.7246,
      "step": 1670
    },
    {
      "epoch": 1.0,
      "step": 1670,
      "total_flos": 1.396344619008e+16,
      "train_loss": 2.9739012581145694,
      "train_runtime": 9187.7213,
      "train_samples_per_second": 11.629,
      "train_steps_per_second": 0.182
    }
  ],
  "logging_steps": 5,
  "max_steps": 1670,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.396344619008e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}