{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 9480,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010548523206751054,
      "grad_norm": 1.2371647357940674,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.5911,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.0724656581878662,
      "learning_rate": 0.00031645569620253165,
      "loss": 7.0366,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8614045977592468,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.3984,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 0.6243458390235901,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.8768,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.4496183395385742,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.5034,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 0.3917015790939331,
      "learning_rate": 0.0009493670886075949,
      "loss": 5.0734,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 0.2791913151741028,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.664,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 0.3657084107398987,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.339,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 0.44374170899391174,
      "learning_rate": 0.0014240506329113926,
      "loss": 4.0971,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 0.5566601753234863,
      "learning_rate": 0.0015,
      "loss": 3.9403,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.3145308792591095,
      "learning_rate": 0.0015,
      "loss": 3.7763,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 0.3487795889377594,
      "learning_rate": 0.0015,
      "loss": 3.6478,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.47254374623298645,
      "learning_rate": 0.0015,
      "loss": 3.5402,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.3688655495643616,
      "learning_rate": 0.0015,
      "loss": 3.4295,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.6684237718582153,
      "learning_rate": 0.0015,
      "loss": 3.3361,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 0.31444546580314636,
      "learning_rate": 0.0015,
      "loss": 3.2668,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 0.6808280944824219,
      "learning_rate": 0.0015,
      "loss": 3.1863,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.35720598697662354,
      "learning_rate": 0.0015,
      "loss": 3.1343,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.47156453132629395,
      "learning_rate": 0.0015,
      "loss": 3.0545,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.39283475279808044,
      "learning_rate": 0.0015,
      "loss": 2.9937,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.413607656955719,
      "learning_rate": 0.0015,
      "loss": 2.931,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.39332082867622375,
      "learning_rate": 0.0015,
      "loss": 2.8887,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.4956569969654083,
      "learning_rate": 0.0015,
      "loss": 2.8105,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.34334513545036316,
      "learning_rate": 0.0015,
      "loss": 2.7606,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 0.4167127311229706,
      "learning_rate": 0.0015,
      "loss": 2.7061,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.3620551824569702,
      "learning_rate": 0.0015,
      "loss": 2.6471,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 0.45425307750701904,
      "learning_rate": 0.0015,
      "loss": 2.61,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.5003123879432678,
      "learning_rate": 0.0015,
      "loss": 2.5594,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.4262688457965851,
      "learning_rate": 0.0015,
      "loss": 2.5234,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 0.3864128291606903,
      "learning_rate": 0.0015,
      "loss": 2.487,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.49673473834991455,
      "learning_rate": 0.0015,
      "loss": 2.4486,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 0.43513303995132446,
      "learning_rate": 0.0015,
      "loss": 2.413,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.41726741194725037,
      "learning_rate": 0.0015,
      "loss": 2.3925,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 0.4520697593688965,
      "learning_rate": 0.0015,
      "loss": 2.3633,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 0.36444756388664246,
      "learning_rate": 0.0015,
      "loss": 2.31,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.45437198877334595,
      "learning_rate": 0.0015,
      "loss": 2.2836,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 0.45183318853378296,
      "learning_rate": 0.0015,
      "loss": 2.2729,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.3789414167404175,
      "learning_rate": 0.0015,
      "loss": 2.2489,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 0.5659546852111816,
      "learning_rate": 0.0015,
      "loss": 2.2257,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.7111340761184692,
      "learning_rate": 0.0015,
      "loss": 2.2001,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 0.3687579333782196,
      "learning_rate": 0.0015,
      "loss": 2.1839,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 0.39448997378349304,
      "learning_rate": 0.0015,
      "loss": 2.1582,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.41239675879478455,
      "learning_rate": 0.0015,
      "loss": 2.1346,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 0.34442898631095886,
      "learning_rate": 0.0015,
      "loss": 2.1145,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.4204893112182617,
      "learning_rate": 0.0015,
      "loss": 2.1103,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.43915101885795593,
      "learning_rate": 0.0015,
      "loss": 2.0917,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.36875662207603455,
      "learning_rate": 0.0015,
      "loss": 2.0652,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.4595106840133667,
      "learning_rate": 0.0015,
      "loss": 2.0615,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.3633001148700714,
      "learning_rate": 0.0015,
      "loss": 2.0351,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.39989691972732544,
      "learning_rate": 0.0015,
      "loss": 2.0241,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 0.38117069005966187,
      "learning_rate": 0.0015,
      "loss": 2.0121,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.36800599098205566,
      "learning_rate": 0.0015,
      "loss": 1.9986,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.3794243335723877,
      "learning_rate": 0.0015,
      "loss": 1.9862,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.5420165061950684,
      "learning_rate": 0.0015,
      "loss": 1.9672,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.421758234500885,
      "learning_rate": 0.0015,
      "loss": 1.9709,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.3748304545879364,
      "learning_rate": 0.0015,
      "loss": 1.9287,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 0.38089877367019653,
      "learning_rate": 0.0015,
      "loss": 1.9324,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 0.6355550289154053,
      "learning_rate": 0.0015,
      "loss": 1.9337,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 0.5314757823944092,
      "learning_rate": 0.0015,
      "loss": 1.9149,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 0.5225743055343628,
      "learning_rate": 0.0015,
      "loss": 1.8945,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 0.40494346618652344,
      "learning_rate": 0.0015,
      "loss": 1.8943,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.46508485078811646,
      "learning_rate": 0.0015,
      "loss": 1.8832,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.4040267765522003,
      "learning_rate": 0.0015,
      "loss": 1.8634,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.4464590549468994,
      "learning_rate": 0.0015,
      "loss": 1.8795,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.3856465518474579,
      "learning_rate": 0.0015,
      "loss": 1.8662,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 0.39368781447410583,
      "learning_rate": 0.0015,
      "loss": 1.8503,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.3902086615562439,
      "learning_rate": 0.0015,
      "loss": 1.8364,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.3802025020122528,
      "learning_rate": 0.0015,
      "loss": 1.8278,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.5012637376785278,
      "learning_rate": 0.0015,
      "loss": 1.8388,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.3923953175544739,
      "learning_rate": 0.0015,
      "loss": 1.8209,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.35562559962272644,
      "learning_rate": 0.0015,
      "loss": 1.8087,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.43287137150764465,
      "learning_rate": 0.0015,
      "loss": 1.7948,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.41877642273902893,
      "learning_rate": 0.0015,
      "loss": 1.7858,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.40727561712265015,
      "learning_rate": 0.0015,
      "loss": 1.7967,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.4644089937210083,
      "learning_rate": 0.0015,
      "loss": 1.7794,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 0.5176643133163452,
      "learning_rate": 0.0015,
      "loss": 1.7787,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.421298086643219,
      "learning_rate": 0.0015,
      "loss": 1.7685,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.3996811509132385,
      "learning_rate": 0.0015,
      "loss": 1.7671,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.37680113315582275,
      "learning_rate": 0.0015,
      "loss": 1.7646,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.4056326746940613,
      "learning_rate": 0.0015,
      "loss": 1.7434,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.45371413230895996,
      "learning_rate": 0.0015,
      "loss": 1.7312,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 0.38185834884643555,
      "learning_rate": 0.0015,
      "loss": 1.7392,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.3576831817626953,
      "learning_rate": 0.0015,
      "loss": 1.7431,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.3754999339580536,
      "learning_rate": 0.0015,
      "loss": 1.7192,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.3862115740776062,
      "learning_rate": 0.0015,
      "loss": 1.7178,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.44392654299736023,
      "learning_rate": 0.0015,
      "loss": 1.7194,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.3966352045536041,
      "learning_rate": 0.0015,
      "loss": 1.7145,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.48476067185401917,
      "learning_rate": 0.0015,
      "loss": 1.7128,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 0.4622691571712494,
      "learning_rate": 0.0015,
      "loss": 1.6989,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 0.4104483723640442,
      "learning_rate": 0.0015,
      "loss": 1.6971,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.39348629117012024,
      "learning_rate": 0.0015,
      "loss": 1.7058,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.4345850646495819,
      "learning_rate": 0.0015,
      "loss": 1.6917,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.3856751620769501,
      "learning_rate": 0.0015,
      "loss": 1.6779,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.3952021896839142,
      "learning_rate": 0.0015,
      "loss": 1.6886,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 0.36113595962524414,
      "learning_rate": 0.0015,
      "loss": 1.6677,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.44029003381729126,
      "learning_rate": 0.0015,
      "loss": 1.674,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 0.42792683839797974,
      "learning_rate": 0.0015,
      "loss": 1.6752,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.39351779222488403,
      "learning_rate": 0.0015,
      "loss": 1.6577,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 0.3800772726535797,
      "learning_rate": 0.0015,
      "loss": 1.6539,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.3503066301345825,
      "learning_rate": 0.0015,
      "loss": 1.6618,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.4740210771560669,
      "learning_rate": 0.0015,
      "loss": 1.6535,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 0.40961724519729614,
      "learning_rate": 0.0015,
      "loss": 1.6449,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 0.38675740361213684,
      "learning_rate": 0.0015,
      "loss": 1.6497,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.3638473451137543,
      "learning_rate": 0.0015,
      "loss": 1.6401,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.346079021692276,
      "learning_rate": 0.0015,
      "loss": 1.6389,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.3508372902870178,
      "learning_rate": 0.0015,
      "loss": 1.6295,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.3539598286151886,
      "learning_rate": 0.0015,
      "loss": 1.6167,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.3764999806880951,
      "learning_rate": 0.0015,
      "loss": 1.6202,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.46206948161125183,
      "learning_rate": 0.0015,
      "loss": 1.6285,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.452450692653656,
      "learning_rate": 0.0015,
      "loss": 1.6173,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.3777218759059906,
      "learning_rate": 0.0015,
      "loss": 1.6223,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.3499910533428192,
      "learning_rate": 0.0015,
      "loss": 1.6123,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 0.5509498715400696,
      "learning_rate": 0.0015,
      "loss": 1.603,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.454050213098526,
      "learning_rate": 0.0015,
      "loss": 1.6144,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 0.5862528085708618,
      "learning_rate": 0.0015,
      "loss": 1.6041,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.3834332525730133,
      "learning_rate": 0.0015,
      "loss": 1.6053,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 0.379902720451355,
      "learning_rate": 0.0015,
      "loss": 1.5993,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.39323538541793823,
      "learning_rate": 0.0015,
      "loss": 1.5846,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.34747618436813354,
      "learning_rate": 0.0015,
      "loss": 1.5844,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.4111934304237366,
      "learning_rate": 0.0015,
      "loss": 1.5977,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.3951559066772461,
      "learning_rate": 0.0015,
      "loss": 1.5886,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.3657253682613373,
      "learning_rate": 0.0015,
      "loss": 1.5738,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.36329299211502075,
      "learning_rate": 0.0015,
      "loss": 1.5659,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.38337430357933044,
      "learning_rate": 0.0015,
      "loss": 1.5815,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.480984628200531,
      "learning_rate": 0.0015,
      "loss": 1.5826,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.37974998354911804,
      "learning_rate": 0.0015,
      "loss": 1.5664,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.3430846333503723,
      "learning_rate": 0.0015,
      "loss": 1.5652,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.3568759560585022,
      "learning_rate": 0.0015,
      "loss": 1.5765,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.3612004518508911,
      "learning_rate": 0.0015,
      "loss": 1.5674,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.34362566471099854,
      "learning_rate": 0.0015,
      "loss": 1.5532,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.36927589774131775,
      "learning_rate": 0.0015,
      "loss": 1.5584,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.40451106429100037,
      "learning_rate": 0.0015,
      "loss": 1.5604,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 0.4525816738605499,
      "learning_rate": 0.0015,
      "loss": 1.5492,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.4472123086452484,
      "learning_rate": 0.0015,
      "loss": 1.5516,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.3997417390346527,
      "learning_rate": 0.0015,
      "loss": 1.555,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.3727695047855377,
      "learning_rate": 0.0015,
      "loss": 1.5489,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.4645366072654724,
      "learning_rate": 0.0015,
      "loss": 1.5449,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.3957315981388092,
      "learning_rate": 0.0015,
      "loss": 1.5384,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 0.39273178577423096,
      "learning_rate": 0.0015,
      "loss": 1.5496,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 0.4525180160999298,
      "learning_rate": 0.0015,
      "loss": 1.5281,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 0.40496838092803955,
      "learning_rate": 0.0015,
      "loss": 1.5339,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.33585745096206665,
      "learning_rate": 0.0015,
      "loss": 1.5359,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.33285295963287354,
      "learning_rate": 0.0015,
      "loss": 1.5294,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.34191590547561646,
      "learning_rate": 0.0015,
      "loss": 1.523,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.3581468164920807,
      "learning_rate": 0.0015,
      "loss": 1.5225,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.3397340476512909,
      "learning_rate": 0.0015,
      "loss": 1.5222,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 0.3838866353034973,
      "learning_rate": 0.0015,
      "loss": 1.5217,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.4081639051437378,
      "learning_rate": 0.0015,
      "loss": 1.5245,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 0.3484359085559845,
      "learning_rate": 0.0015,
      "loss": 1.5237,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.3435753285884857,
      "learning_rate": 0.0015,
      "loss": 1.5168,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.3409862518310547,
      "learning_rate": 0.0015,
      "loss": 1.5182,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.35892271995544434,
      "learning_rate": 0.0015,
      "loss": 1.5208,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.3690599501132965,
      "learning_rate": 0.0015,
      "loss": 1.5023,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.3629121482372284,
      "learning_rate": 0.0015,
      "loss": 1.4942,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 0.3664407730102539,
      "learning_rate": 0.0015,
      "loss": 1.4971,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.40004992485046387,
      "learning_rate": 0.0015,
      "loss": 1.5002,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.3557882308959961,
      "learning_rate": 0.0015,
      "loss": 1.4924,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.3498414158821106,
      "learning_rate": 0.0015,
      "loss": 1.5023,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 0.369983971118927,
      "learning_rate": 0.0015,
      "loss": 1.5012,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.3442850410938263,
      "learning_rate": 0.0015,
      "loss": 1.5014,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.4010471701622009,
      "learning_rate": 0.0015,
      "loss": 1.4968,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.34040263295173645,
      "learning_rate": 0.0015,
      "loss": 1.4674,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.3703126013278961,
      "learning_rate": 0.0015,
      "loss": 1.4828,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.34696683287620544,
      "learning_rate": 0.0015,
      "loss": 1.4879,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.35845789313316345,
      "learning_rate": 0.0015,
      "loss": 1.4797,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.3321089744567871,
      "learning_rate": 0.0015,
      "loss": 1.4759,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.39269906282424927,
      "learning_rate": 0.0015,
      "loss": 1.4879,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 0.46313339471817017,
      "learning_rate": 0.0015,
      "loss": 1.4745,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.4027881324291229,
      "learning_rate": 0.0015,
      "loss": 1.4694,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.4224802851676941,
      "learning_rate": 0.0015,
      "loss": 1.4858,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.3397369384765625,
      "learning_rate": 0.0015,
      "loss": 1.461,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.3592316508293152,
      "learning_rate": 0.0015,
      "loss": 1.4617,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.3517380356788635,
      "learning_rate": 0.0015,
      "loss": 1.4641,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.37348607182502747,
      "learning_rate": 0.0015,
      "loss": 1.467,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.3547249138355255,
      "learning_rate": 0.0015,
      "loss": 1.4893,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.34048789739608765,
      "learning_rate": 0.0015,
      "loss": 1.4669,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.3326258063316345,
      "learning_rate": 0.0015,
      "loss": 1.4703,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.3598375618457794,
      "learning_rate": 0.0015,
      "loss": 1.4683,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 0.35684293508529663,
      "learning_rate": 0.0015,
      "loss": 1.4708,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.3906477689743042,
      "learning_rate": 0.0015,
      "loss": 1.4657,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.37610185146331787,
      "learning_rate": 0.0015,
      "loss": 1.454,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.37533876299858093,
      "learning_rate": 0.0015,
      "loss": 1.4522,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.3618168830871582,
      "learning_rate": 0.0015,
      "loss": 1.453,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.3646780550479889,
      "learning_rate": 0.0015,
      "loss": 1.4548,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.3709734380245209,
      "learning_rate": 0.0015,
      "loss": 1.4526,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.36328479647636414,
      "learning_rate": 0.0015,
      "loss": 1.4575,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.35170361399650574,
      "learning_rate": 0.0015,
      "loss": 1.4564,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.3583771586418152,
      "learning_rate": 0.0015,
      "loss": 1.4424,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.3604058623313904,
      "learning_rate": 0.0015,
      "loss": 1.4509,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 0.38244473934173584,
      "learning_rate": 0.0015,
      "loss": 1.4478,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 0.47545963525772095,
      "learning_rate": 0.0015,
      "loss": 1.4484,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.37117186188697815,
      "learning_rate": 0.0015,
      "loss": 1.4483,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 0.3921581208705902,
      "learning_rate": 0.0015,
      "loss": 1.4501,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.3382081985473633,
      "learning_rate": 0.0015,
      "loss": 1.439,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.364330530166626,
      "learning_rate": 0.0015,
      "loss": 1.4374,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.3328717052936554,
      "learning_rate": 0.0015,
      "loss": 1.4362,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.3981072008609772,
      "learning_rate": 0.0015,
      "loss": 1.447,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.3473067283630371,
      "learning_rate": 0.0015,
      "loss": 1.4267,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.3436720073223114,
      "learning_rate": 0.0015,
      "loss": 1.428,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.3526890277862549,
      "learning_rate": 0.0015,
      "loss": 1.4286,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.364200621843338,
      "learning_rate": 0.0015,
      "loss": 1.4363,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 0.4152698814868927,
      "learning_rate": 0.0015,
      "loss": 1.448,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.4520832896232605,
      "learning_rate": 0.0015,
      "loss": 1.4318,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.39879027009010315,
      "learning_rate": 0.0015,
      "loss": 1.4327,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.3920576274394989,
      "learning_rate": 0.0015,
      "loss": 1.4296,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.33325308561325073,
      "learning_rate": 0.0015,
      "loss": 1.4205,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 0.41690853238105774,
      "learning_rate": 0.0015,
      "loss": 1.426,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.37545451521873474,
      "learning_rate": 0.0015,
      "loss": 1.4286,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.3608053922653198,
      "learning_rate": 0.0015,
      "loss": 1.4183,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.3502177596092224,
      "learning_rate": 0.0015,
      "loss": 1.4256,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.37644052505493164,
      "learning_rate": 0.0015,
      "loss": 1.4128,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.3824097514152527,
      "learning_rate": 0.0015,
      "loss": 1.4146,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.3496038317680359,
      "learning_rate": 0.0015,
      "loss": 1.4162,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.3546043038368225,
      "learning_rate": 0.0015,
      "loss": 1.4142,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.37273067235946655,
      "learning_rate": 0.0015,
      "loss": 1.4209,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 0.46220526099205017,
      "learning_rate": 0.0015,
      "loss": 1.4323,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.3861055374145508,
      "learning_rate": 0.0015,
      "loss": 1.4159,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.35327857732772827,
      "learning_rate": 0.0015,
      "loss": 1.4006,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.49925553798675537,
      "learning_rate": 0.0015,
      "loss": 1.4182,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.4116460978984833,
      "learning_rate": 0.0015,
      "loss": 1.4206,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.3701082170009613,
      "learning_rate": 0.0015,
      "loss": 1.4007,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.4048565924167633,
      "learning_rate": 0.0015,
      "loss": 1.4166,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 0.34562838077545166,
      "learning_rate": 0.0015,
      "loss": 1.4059,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.37388190627098083,
      "learning_rate": 0.0015,
      "loss": 1.4033,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.3642711341381073,
      "learning_rate": 0.0015,
      "loss": 1.4046,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.4138195514678955,
      "learning_rate": 0.0015,
      "loss": 1.3919,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.3694874048233032,
      "learning_rate": 0.0015,
      "loss": 1.4251,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.33054301142692566,
      "learning_rate": 0.0015,
      "loss": 1.4104,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.35614264011383057,
      "learning_rate": 0.0015,
      "loss": 1.3991,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.3298373520374298,
      "learning_rate": 0.0015,
      "loss": 1.3965,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.3741556704044342,
      "learning_rate": 0.0015,
      "loss": 1.3957,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.3534122705459595,
      "learning_rate": 0.0015,
      "loss": 1.4096,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.3374796211719513,
      "learning_rate": 0.0015,
      "loss": 1.3925,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.3883809745311737,
      "learning_rate": 0.0015,
      "loss": 1.3853,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.4112820625305176,
      "learning_rate": 0.0015,
      "loss": 1.3886,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.4407918155193329,
      "learning_rate": 0.0015,
      "loss": 1.4009,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.38165411353111267,
      "learning_rate": 0.0015,
      "loss": 1.3963,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.3438486158847809,
      "learning_rate": 0.0015,
      "loss": 1.3921,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.38707923889160156,
      "learning_rate": 0.0015,
      "loss": 1.3906,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.35939821600914,
      "learning_rate": 0.0015,
      "loss": 1.3911,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.3938080072402954,
      "learning_rate": 0.0015,
      "loss": 1.3951,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.37135642766952515,
      "learning_rate": 0.0015,
      "loss": 1.3871,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.41192826628685,
      "learning_rate": 0.0015,
      "loss": 1.3839,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.37276938557624817,
      "learning_rate": 0.0015,
      "loss": 1.3946,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.3469187915325165,
      "learning_rate": 0.0015,
      "loss": 1.3788,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.3916131556034088,
      "learning_rate": 0.0015,
      "loss": 1.3964,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.35671016573905945,
      "learning_rate": 0.0015,
      "loss": 1.3874,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.37222930788993835,
      "learning_rate": 0.0015,
      "loss": 1.3916,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.3410083055496216,
      "learning_rate": 0.0015,
      "loss": 1.3775,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.39879265427589417,
      "learning_rate": 0.0015,
      "loss": 1.3804,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.3476262092590332,
      "learning_rate": 0.0015,
      "loss": 1.3803,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.38001757860183716,
      "learning_rate": 0.0015,
      "loss": 1.3861,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.39220911264419556,
      "learning_rate": 0.0015,
      "loss": 1.3812,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 0.4335632026195526,
      "learning_rate": 0.0015,
      "loss": 1.3804,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.3345543444156647,
      "learning_rate": 0.0015,
      "loss": 1.3789,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.3646849989891052,
      "learning_rate": 0.0015,
      "loss": 1.3853,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.3379766345024109,
      "learning_rate": 0.0015,
      "loss": 1.3798,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.3333427309989929,
      "learning_rate": 0.0015,
      "loss": 1.3782,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.3338565528392792,
      "learning_rate": 0.0015,
      "loss": 1.3819,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.3742698132991791,
      "learning_rate": 0.0015,
      "loss": 1.3804,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.38040274381637573,
      "learning_rate": 0.0015,
      "loss": 1.3784,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.4038099944591522,
      "learning_rate": 0.0015,
      "loss": 1.3782,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.3927282691001892,
      "learning_rate": 0.0015,
      "loss": 1.3669,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.39924538135528564,
      "learning_rate": 0.0015,
      "loss": 1.3796,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.35913312435150146,
      "learning_rate": 0.0015,
      "loss": 1.3726,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.3561171293258667,
      "learning_rate": 0.0015,
      "loss": 1.3661,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.3356326222419739,
      "learning_rate": 0.0015,
      "loss": 1.3696,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.3672836422920227,
      "learning_rate": 0.0015,
      "loss": 1.3672,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.3602263629436493,
      "learning_rate": 0.0015,
      "loss": 1.3699,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.37418046593666077,
      "learning_rate": 0.0015,
      "loss": 1.3623,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.3903089463710785,
      "learning_rate": 0.0015,
      "loss": 1.3716,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.361920565366745,
      "learning_rate": 0.0015,
      "loss": 1.3704,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.33929145336151123,
      "learning_rate": 0.0015,
      "loss": 1.3527,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 0.3812505006790161,
      "learning_rate": 0.0015,
      "loss": 1.3633,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.348438024520874,
      "learning_rate": 0.0015,
      "loss": 1.3679,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.3531169593334198,
      "learning_rate": 0.0015,
      "loss": 1.3743,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.34419044852256775,
      "learning_rate": 0.0015,
      "loss": 1.372,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.3198499381542206,
      "learning_rate": 0.0015,
      "loss": 1.3674,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.32899436354637146,
      "learning_rate": 0.0015,
      "loss": 1.3651,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.36924058198928833,
      "learning_rate": 0.0015,
      "loss": 1.3639,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.3184834122657776,
      "learning_rate": 0.0015,
      "loss": 1.3648,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.4172210097312927,
      "learning_rate": 0.0015,
      "loss": 1.3676,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 0.3834500312805176,
      "learning_rate": 0.0015,
      "loss": 1.3455,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 0.3521396517753601,
      "learning_rate": 0.0015,
      "loss": 1.3502,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.31845158338546753,
      "learning_rate": 0.0015,
      "loss": 1.3533,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.3387175500392914,
      "learning_rate": 0.0015,
      "loss": 1.3583,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.36378470063209534,
      "learning_rate": 0.0015,
      "loss": 1.3588,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.3735678791999817,
      "learning_rate": 0.0015,
      "loss": 1.3471,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.3498466908931732,
      "learning_rate": 0.0015,
      "loss": 1.3488,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.3306233882904053,
      "learning_rate": 0.0015,
      "loss": 1.3528,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.35176217555999756,
      "learning_rate": 0.0015,
      "loss": 1.3582,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.36599618196487427,
      "learning_rate": 0.0015,
      "loss": 1.3515,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.3750476539134979,
      "learning_rate": 0.0015,
      "loss": 1.3497,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.36336538195610046,
      "learning_rate": 0.0015,
      "loss": 1.3542,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.36796873807907104,
      "learning_rate": 0.0015,
      "loss": 1.3614,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.4234572649002075,
      "learning_rate": 0.0015,
      "loss": 1.3459,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 0.3786514401435852,
      "learning_rate": 0.0015,
      "loss": 1.3409,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.3337559401988983,
      "learning_rate": 0.0015,
      "loss": 1.3446,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.3283361792564392,
      "learning_rate": 0.0015,
      "loss": 1.3599,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.3510412871837616,
      "learning_rate": 0.0015,
      "loss": 1.3363,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.34402918815612793,
      "learning_rate": 0.0015,
      "loss": 1.3446,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.34152328968048096,
      "learning_rate": 0.0015,
      "loss": 1.3479,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.3709448575973511,
      "learning_rate": 0.0015,
      "loss": 1.3391,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.4077426791191101,
      "learning_rate": 0.0015,
      "loss": 1.3466,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.3553972840309143,
      "learning_rate": 0.0015,
      "loss": 1.3401,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.32798343896865845,
      "learning_rate": 0.0015,
      "loss": 1.3504,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.3771567642688751,
      "learning_rate": 0.0015,
      "loss": 1.3409,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.3586834967136383,
      "learning_rate": 0.0015,
      "loss": 1.3315,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.32795462012290955,
      "learning_rate": 0.0015,
      "loss": 1.3485,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.36836713552474976,
      "learning_rate": 0.0015,
      "loss": 1.3378,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.41258755326271057,
      "learning_rate": 0.0015,
      "loss": 1.3385,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.3700130879878998,
      "learning_rate": 0.0015,
      "loss": 1.3392,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.35686036944389343,
      "learning_rate": 0.0015,
      "loss": 1.3516,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.4040427505970001,
      "learning_rate": 0.0015,
      "loss": 1.3291,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.3243670165538788,
      "learning_rate": 0.0015,
      "loss": 1.3382,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.3412677049636841,
      "learning_rate": 0.0015,
      "loss": 1.3403,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.3531607985496521,
      "learning_rate": 0.0015,
      "loss": 1.3417,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.3264332115650177,
      "learning_rate": 0.0015,
      "loss": 1.3345,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.38159504532814026,
      "learning_rate": 0.0015,
      "loss": 1.3335,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.34057164192199707,
      "learning_rate": 0.0015,
      "loss": 1.3303,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.4352438151836395,
      "learning_rate": 0.0015,
      "loss": 1.3452,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.3315022587776184,
      "learning_rate": 0.0015,
      "loss": 1.3366,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.33373403549194336,
      "learning_rate": 0.0015,
      "loss": 1.3275,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.33522918820381165,
      "learning_rate": 0.0015,
      "loss": 1.3095,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.32375264167785645,
      "learning_rate": 0.0015,
      "loss": 1.3337,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.33494749665260315,
      "learning_rate": 0.0015,
      "loss": 1.3273,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.34836721420288086,
      "learning_rate": 0.0015,
      "loss": 1.3428,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.32541200518608093,
      "learning_rate": 0.0015,
      "loss": 1.3393,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 0.3796813189983368,
      "learning_rate": 0.0015,
      "loss": 1.3199,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.3402756154537201,
      "learning_rate": 0.0015,
      "loss": 1.3315,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.34173041582107544,
      "learning_rate": 0.0015,
      "loss": 1.3347,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.33362719416618347,
      "learning_rate": 0.0015,
      "loss": 1.3397,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.345304012298584,
      "learning_rate": 0.0015,
      "loss": 1.3266,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.36389777064323425,
      "learning_rate": 0.0015,
      "loss": 1.3297,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.37752678990364075,
      "learning_rate": 0.0015,
      "loss": 1.3234,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.33431223034858704,
      "learning_rate": 0.0015,
      "loss": 1.3252,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.4039674699306488,
      "learning_rate": 0.0015,
      "loss": 1.3316,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.3299323320388794,
      "learning_rate": 0.0015,
      "loss": 1.3242,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.3494005799293518,
      "learning_rate": 0.0015,
      "loss": 1.3218,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.3371613025665283,
      "learning_rate": 0.0015,
      "loss": 1.3285,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.3141007125377655,
      "learning_rate": 0.0015,
      "loss": 1.3247,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.3440660536289215,
      "learning_rate": 0.0015,
      "loss": 1.3065,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.3393572270870209,
      "learning_rate": 0.0015,
      "loss": 1.3221,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.35718950629234314,
      "learning_rate": 0.0015,
      "loss": 1.3221,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.41040876507759094,
      "learning_rate": 0.0015,
      "loss": 1.3196,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.35438480973243713,
      "learning_rate": 0.0015,
      "loss": 1.3131,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.35374361276626587,
      "learning_rate": 0.0015,
      "loss": 1.3287,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.36048999428749084,
      "learning_rate": 0.0015,
      "loss": 1.3339,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.3104354441165924,
      "learning_rate": 0.0015,
      "loss": 1.3235,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.3356916606426239,
      "learning_rate": 0.0015,
      "loss": 1.3059,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.3305603563785553,
      "learning_rate": 0.0015,
      "loss": 1.3261,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.35006171464920044,
      "learning_rate": 0.0015,
      "loss": 1.3231,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.3865632116794586,
      "learning_rate": 0.0015,
      "loss": 1.3328,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 0.3990705609321594,
      "learning_rate": 0.0015,
      "loss": 1.3248,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.32694146037101746,
      "learning_rate": 0.0015,
      "loss": 1.3134,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.36591851711273193,
      "learning_rate": 0.0015,
      "loss": 1.3184,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.33968791365623474,
      "learning_rate": 0.0015,
      "loss": 1.3137,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.4435545802116394,
      "learning_rate": 0.0015,
      "loss": 1.3159,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.3401292562484741,
      "learning_rate": 0.0015,
      "loss": 1.326,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.3403913378715515,
      "learning_rate": 0.0015,
      "loss": 1.3122,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.3234756290912628,
      "learning_rate": 0.0015,
      "loss": 1.3101,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.33335044980049133,
      "learning_rate": 0.0015,
      "loss": 1.3139,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.33656445145606995,
      "learning_rate": 0.0015,
      "loss": 1.3098,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.39324823021888733,
      "learning_rate": 0.0015,
      "loss": 1.3116,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.3268013894557953,
      "learning_rate": 0.0015,
      "loss": 1.3187,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.34896695613861084,
      "learning_rate": 0.0015,
      "loss": 1.3125,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.34891918301582336,
      "learning_rate": 0.0015,
      "loss": 1.3068,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.48617294430732727,
      "learning_rate": 0.0015,
      "loss": 1.3129,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.3484684228897095,
      "learning_rate": 0.0015,
      "loss": 1.3198,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.40289464592933655,
      "learning_rate": 0.0015,
      "loss": 1.3212,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.3295171856880188,
      "learning_rate": 0.0015,
      "loss": 1.3115,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.4097455143928528,
      "learning_rate": 0.0015,
      "loss": 1.2935,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.34945106506347656,
      "learning_rate": 0.0015,
      "loss": 1.3109,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.368396520614624,
      "learning_rate": 0.0015,
      "loss": 1.3092,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.39238911867141724,
      "learning_rate": 0.0015,
      "loss": 1.3136,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.3448035717010498,
      "learning_rate": 0.0015,
      "loss": 1.3017,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.3241119980812073,
      "learning_rate": 0.0015,
      "loss": 1.3197,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.3216939866542816,
      "learning_rate": 0.0015,
      "loss": 1.3042,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.4088415503501892,
      "learning_rate": 0.0015,
      "loss": 1.3082,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.3490767478942871,
      "learning_rate": 0.0015,
      "loss": 1.3148,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.34655386209487915,
      "learning_rate": 0.0015,
      "loss": 1.3002,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.31820032000541687,
      "learning_rate": 0.0015,
      "loss": 1.2961,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.344877690076828,
      "learning_rate": 0.0015,
      "loss": 1.3002,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.3609689176082611,
      "learning_rate": 0.0015,
      "loss": 1.3139,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.36878663301467896,
      "learning_rate": 0.0015,
      "loss": 1.3012,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.36532774567604065,
      "learning_rate": 0.0015,
      "loss": 1.3117,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.3955930769443512,
      "learning_rate": 0.0015,
      "loss": 1.3018,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.331284761428833,
      "learning_rate": 0.0015,
      "loss": 1.293,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.3166908323764801,
      "learning_rate": 0.0015,
      "loss": 1.2926,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.4257475733757019,
      "learning_rate": 0.0015,
      "loss": 1.2952,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.3384430706501007,
      "learning_rate": 0.0015,
      "loss": 1.3007,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.3213377892971039,
      "learning_rate": 0.0015,
      "loss": 1.2939,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.33240780234336853,
      "learning_rate": 0.0015,
      "loss": 1.2996,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.3315512239933014,
      "learning_rate": 0.0015,
      "loss": 1.2951,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.31810611486434937,
      "learning_rate": 0.0015,
      "loss": 1.2988,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.3297857940196991,
      "learning_rate": 0.0015,
      "loss": 1.3009,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.43589648604393005,
      "learning_rate": 0.0015,
      "loss": 1.2905,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.33363038301467896,
      "learning_rate": 0.0015,
      "loss": 1.3039,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.3441077172756195,
      "learning_rate": 0.0015,
      "loss": 1.2928,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.318351149559021,
      "learning_rate": 0.0015,
      "loss": 1.3061,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.367177814245224,
      "learning_rate": 0.0015,
      "loss": 1.295,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.3348611891269684,
      "learning_rate": 0.0015,
      "loss": 1.2949,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.32858869433403015,
      "learning_rate": 0.0015,
      "loss": 1.2884,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.32444247603416443,
      "learning_rate": 0.0015,
      "loss": 1.2979,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.34852734208106995,
      "learning_rate": 0.0015,
      "loss": 1.3015,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.33892375230789185,
      "learning_rate": 0.0015,
      "loss": 1.2994,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.3241480886936188,
      "learning_rate": 0.0015,
      "loss": 1.2973,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.4842641353607178,
      "learning_rate": 0.0015,
      "loss": 1.3032,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.3711332082748413,
      "learning_rate": 0.0015,
      "loss": 1.2966,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.3361760377883911,
      "learning_rate": 0.0015,
      "loss": 1.2948,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.3823705315589905,
      "learning_rate": 0.0015,
      "loss": 1.2865,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.33324065804481506,
      "learning_rate": 0.0015,
      "loss": 1.2931,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.3460291922092438,
      "learning_rate": 0.0015,
      "loss": 1.2852,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.3603265583515167,
      "learning_rate": 0.0015,
      "loss": 1.2882,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.3887493312358856,
      "learning_rate": 0.0015,
      "loss": 1.2957,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.3834715783596039,
      "learning_rate": 0.0015,
      "loss": 1.288,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.375080943107605,
      "learning_rate": 0.0015,
      "loss": 1.2877,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.32457664608955383,
      "learning_rate": 0.0015,
      "loss": 1.2849,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.3321995437145233,
      "learning_rate": 0.0015,
      "loss": 1.2886,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.3851778209209442,
      "learning_rate": 0.0015,
      "loss": 1.2932,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.3327183425426483,
      "learning_rate": 0.0015,
      "loss": 1.2827,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.315941721200943,
      "learning_rate": 0.0015,
      "loss": 1.2908,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 0.3883848488330841,
      "learning_rate": 0.0015,
      "loss": 1.2886,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.35884931683540344,
      "learning_rate": 0.0015,
      "loss": 1.2842,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.41224420070648193,
      "learning_rate": 0.0015,
      "loss": 1.2897,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 0.34621503949165344,
      "learning_rate": 0.0015,
      "loss": 1.2967,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.310474157333374,
      "learning_rate": 0.0015,
      "loss": 1.2854,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.36023226380348206,
      "learning_rate": 0.0015,
      "loss": 1.2778,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.33774444460868835,
      "learning_rate": 0.0015,
      "loss": 1.2888,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.3492637574672699,
      "learning_rate": 0.0015,
      "loss": 1.2899,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.3130638301372528,
      "learning_rate": 0.0015,
      "loss": 1.2858,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.34976881742477417,
      "learning_rate": 0.0015,
      "loss": 1.2814,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.3387364149093628,
      "learning_rate": 0.0015,
      "loss": 1.2885,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.37869808077812195,
      "learning_rate": 0.0015,
      "loss": 1.2848,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.367775559425354,
      "learning_rate": 0.0015,
      "loss": 1.2837,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.35876554250717163,
      "learning_rate": 0.0015,
      "loss": 1.2969,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.33714690804481506,
      "learning_rate": 0.0015,
      "loss": 1.2769,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.43783363699913025,
      "learning_rate": 0.0015,
      "loss": 1.2823,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.3282812535762787,
      "learning_rate": 0.0015,
      "loss": 1.2804,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.314325213432312,
      "learning_rate": 0.0015,
      "loss": 1.2849,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.3292233347892761,
      "learning_rate": 0.0015,
      "loss": 1.2959,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.33296987414360046,
      "learning_rate": 0.0015,
      "loss": 1.281,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.35254812240600586,
      "learning_rate": 0.0015,
      "loss": 1.2845,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.3571605980396271,
      "learning_rate": 0.0015,
      "loss": 1.2818,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.3396008312702179,
      "learning_rate": 0.0015,
      "loss": 1.2875,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.3483985662460327,
      "learning_rate": 0.0015,
      "loss": 1.275,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.36943528056144714,
      "learning_rate": 0.0015,
      "loss": 1.2838,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.3278040885925293,
      "learning_rate": 0.0015,
      "loss": 1.2895,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.34484514594078064,
      "learning_rate": 0.0015,
      "loss": 1.2773,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.4180009365081787,
      "learning_rate": 0.0015,
      "loss": 1.2789,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.3349873125553131,
      "learning_rate": 0.0015,
      "loss": 1.2832,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.32025009393692017,
      "learning_rate": 0.0015,
      "loss": 1.2658,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.35468825697898865,
      "learning_rate": 0.0015,
      "loss": 1.2846,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.338003545999527,
      "learning_rate": 0.0015,
      "loss": 1.2838,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.41642409563064575,
      "learning_rate": 0.0015,
      "loss": 1.2839,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.3109353184700012,
      "learning_rate": 0.0015,
      "loss": 1.2749,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.35001808404922485,
      "learning_rate": 0.0015,
      "loss": 1.2547,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.3922937512397766,
      "learning_rate": 0.0015,
      "loss": 1.2803,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.3586471974849701,
      "learning_rate": 0.0015,
      "loss": 1.2677,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.34954914450645447,
      "learning_rate": 0.0015,
      "loss": 1.2831,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 0.3975285291671753,
      "learning_rate": 0.0015,
      "loss": 1.2866,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.33980458974838257,
      "learning_rate": 0.0015,
      "loss": 1.2652,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.3447878360748291,
      "learning_rate": 0.0015,
      "loss": 1.2797,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.3304397165775299,
      "learning_rate": 0.0015,
      "loss": 1.2827,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.33216533064842224,
      "learning_rate": 0.0015,
      "loss": 1.2829,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.3622681200504303,
      "learning_rate": 0.0015,
      "loss": 1.2697,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.3417949676513672,
      "learning_rate": 0.0015,
      "loss": 1.2715,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.30141401290893555,
      "learning_rate": 0.0015,
      "loss": 1.2669,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.3838488459587097,
      "learning_rate": 0.0015,
      "loss": 1.2673,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.3914671838283539,
      "learning_rate": 0.0015,
      "loss": 1.275,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.34327882528305054,
      "learning_rate": 0.0015,
      "loss": 1.2759,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.41521546244621277,
      "learning_rate": 0.0015,
      "loss": 1.2751,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.3268080949783325,
      "learning_rate": 0.0015,
      "loss": 1.2678,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.34257772564888,
      "learning_rate": 0.0015,
      "loss": 1.2655,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.3677382171154022,
      "learning_rate": 0.0015,
      "loss": 1.2863,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.3203747868537903,
      "learning_rate": 0.0015,
      "loss": 1.2714,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.30679064989089966,
      "learning_rate": 0.0015,
      "loss": 1.2727,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.31779491901397705,
      "learning_rate": 0.0015,
      "loss": 1.2805,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.4639074504375458,
      "learning_rate": 0.0015,
      "loss": 1.2648,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.35582196712493896,
      "learning_rate": 0.0015,
      "loss": 1.2671,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.32101449370384216,
      "learning_rate": 0.0015,
      "loss": 1.2664,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.31420451402664185,
      "learning_rate": 0.0015,
      "loss": 1.2863,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.3510427474975586,
      "learning_rate": 0.0015,
      "loss": 1.2685,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.3493404984474182,
      "learning_rate": 0.0015,
      "loss": 1.2643,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.33023601770401,
      "learning_rate": 0.0015,
      "loss": 1.2788,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.33449527621269226,
      "learning_rate": 0.0015,
      "loss": 1.2657,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.3172394037246704,
      "learning_rate": 0.0015,
      "loss": 1.258,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.3701837956905365,
      "learning_rate": 0.0015,
      "loss": 1.2801,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.34527572989463806,
      "learning_rate": 0.0015,
      "loss": 1.2679,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.3702089786529541,
      "learning_rate": 0.0015,
      "loss": 1.2764,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.36963438987731934,
      "learning_rate": 0.0015,
      "loss": 1.2698,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.3635721206665039,
      "learning_rate": 0.0015,
      "loss": 1.2623,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.3239406943321228,
      "learning_rate": 0.0015,
      "loss": 1.2705,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.34057164192199707,
      "learning_rate": 0.0015,
      "loss": 1.2689,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.35554683208465576,
      "learning_rate": 0.0015,
      "loss": 1.276,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.3254257142543793,
      "learning_rate": 0.0015,
      "loss": 1.258,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.36377695202827454,
      "learning_rate": 0.0015,
      "loss": 1.2638,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.3804371654987335,
      "learning_rate": 0.0015,
      "loss": 1.2603,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.34835729002952576,
      "learning_rate": 0.0015,
      "loss": 1.2637,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.333615779876709,
      "learning_rate": 0.0015,
      "loss": 1.273,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.3416522741317749,
      "learning_rate": 0.0015,
      "loss": 1.2677,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.33487850427627563,
      "learning_rate": 0.0015,
      "loss": 1.2629,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.32856062054634094,
      "learning_rate": 0.0015,
      "loss": 1.2473,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.32477837800979614,
      "learning_rate": 0.0015,
      "loss": 1.2735,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.31196415424346924,
      "learning_rate": 0.0015,
      "loss": 1.2604,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.3626239001750946,
      "learning_rate": 0.0015,
      "loss": 1.2754,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.347139835357666,
      "learning_rate": 0.0015,
      "loss": 1.2616,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.332815557718277,
      "learning_rate": 0.0015,
      "loss": 1.2756,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.32809755206108093,
      "learning_rate": 0.0015,
      "loss": 1.2651,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.35669374465942383,
      "learning_rate": 0.0015,
      "loss": 1.2478,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.3932604193687439,
      "learning_rate": 0.0015,
      "loss": 1.2627,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.3809468746185303,
      "learning_rate": 0.0015,
      "loss": 1.2614,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.3322828710079193,
      "learning_rate": 0.0015,
      "loss": 1.2611,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.30540966987609863,
      "learning_rate": 0.0015,
      "loss": 1.2751,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.3536911904811859,
      "learning_rate": 0.0015,
      "loss": 1.2687,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.37020963430404663,
      "learning_rate": 0.0015,
      "loss": 1.2653,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.3286992013454437,
      "learning_rate": 0.0015,
      "loss": 1.2642,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.4192366600036621,
      "learning_rate": 0.0015,
      "loss": 1.2745,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.32625076174736023,
      "learning_rate": 0.0015,
      "loss": 1.2572,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.3306078314781189,
      "learning_rate": 0.0015,
      "loss": 1.2675,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.33644092082977295,
      "learning_rate": 0.0015,
      "loss": 1.2556,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.3233281373977661,
      "learning_rate": 0.0015,
      "loss": 1.2636,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.34870925545692444,
      "learning_rate": 0.0015,
      "loss": 1.2567,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.3786031901836395,
      "learning_rate": 0.0015,
      "loss": 1.2648,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.3718455731868744,
      "learning_rate": 0.0015,
      "loss": 1.2562,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.40321359038352966,
      "learning_rate": 0.0015,
      "loss": 1.263,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.44219493865966797,
      "learning_rate": 0.0015,
      "loss": 1.2704,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.326164186000824,
      "learning_rate": 0.0015,
      "loss": 1.2592,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.34893083572387695,
      "learning_rate": 0.0015,
      "loss": 1.2584,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.35736313462257385,
      "learning_rate": 0.0015,
      "loss": 1.2661,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.39100855588912964,
      "learning_rate": 0.0015,
      "loss": 1.2524,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.32789599895477295,
      "learning_rate": 0.0015,
      "loss": 1.2548,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.33129817247390747,
      "learning_rate": 0.0015,
      "loss": 1.2625,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.32572874426841736,
      "learning_rate": 0.0015,
      "loss": 1.2548,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.3237755000591278,
      "learning_rate": 0.0015,
      "loss": 1.2515,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.355275958776474,
      "learning_rate": 0.0015,
      "loss": 1.257,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.3404082953929901,
      "learning_rate": 0.0015,
      "loss": 1.2593,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.3304276466369629,
      "learning_rate": 0.0015,
      "loss": 1.2517,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.33989864587783813,
      "learning_rate": 0.0015,
      "loss": 1.2522,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.341671884059906,
      "learning_rate": 0.0015,
      "loss": 1.2519,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.3551168143749237,
      "learning_rate": 0.0015,
      "loss": 1.2572,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.346642404794693,
      "learning_rate": 0.0015,
      "loss": 1.2501,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.35103023052215576,
      "learning_rate": 0.0015,
      "loss": 1.2483,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.3185129761695862,
      "learning_rate": 0.0015,
      "loss": 1.2482,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.3302600383758545,
      "learning_rate": 0.0015,
      "loss": 1.2504,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.32768237590789795,
      "learning_rate": 0.0015,
      "loss": 1.2538,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.35727646946907043,
      "learning_rate": 0.0015,
      "loss": 1.254,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.367353230714798,
      "learning_rate": 0.0015,
      "loss": 1.2548,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.3531593978404999,
      "learning_rate": 0.0015,
      "loss": 1.2631,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.32804006338119507,
      "learning_rate": 0.0015,
      "loss": 1.251,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.31846901774406433,
      "learning_rate": 0.0015,
      "loss": 1.2583,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.36841240525245667,
      "learning_rate": 0.0015,
      "loss": 1.253,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.46265220642089844,
      "learning_rate": 0.0015,
      "loss": 1.2458,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.32472917437553406,
      "learning_rate": 0.0015,
      "loss": 1.2434,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.33125030994415283,
      "learning_rate": 0.0015,
      "loss": 1.2626,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.3427261710166931,
      "learning_rate": 0.0015,
      "loss": 1.265,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.32336708903312683,
      "learning_rate": 0.0015,
      "loss": 1.2505,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.31386277079582214,
      "learning_rate": 0.0015,
      "loss": 1.2485,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.32289138436317444,
      "learning_rate": 0.0015,
      "loss": 1.2529,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.33690229058265686,
      "learning_rate": 0.0015,
      "loss": 1.2475,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.36602291464805603,
      "learning_rate": 0.0015,
      "loss": 1.2519,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.35524094104766846,
      "learning_rate": 0.0015,
      "loss": 1.2586,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.3474019467830658,
      "learning_rate": 0.0015,
      "loss": 1.2555,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.3641188144683838,
      "learning_rate": 0.0015,
      "loss": 1.2621,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.325612872838974,
      "learning_rate": 0.0015,
      "loss": 1.2555,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.4651053845882416,
      "learning_rate": 0.0015,
      "loss": 1.2517,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.32629069685935974,
      "learning_rate": 0.0015,
      "loss": 1.2581,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.3221701681613922,
      "learning_rate": 0.0015,
      "loss": 1.2586,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.33370065689086914,
      "learning_rate": 0.0015,
      "loss": 1.2393,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.35515671968460083,
      "learning_rate": 0.0015,
      "loss": 1.2423,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.36871209740638733,
      "learning_rate": 0.0015,
      "loss": 1.249,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.3117012083530426,
      "learning_rate": 0.0015,
      "loss": 1.2526,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.3552721440792084,
      "learning_rate": 0.0015,
      "loss": 1.2502,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.33427029848098755,
      "learning_rate": 0.0015,
      "loss": 1.2486,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.34898853302001953,
      "learning_rate": 0.0015,
      "loss": 1.2244,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.33162492513656616,
      "learning_rate": 0.0015,
      "loss": 1.2322,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.36996203660964966,
      "learning_rate": 0.0015,
      "loss": 1.2348,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.30883604288101196,
      "learning_rate": 0.0015,
      "loss": 1.2451,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.3355100750923157,
      "learning_rate": 0.0015,
      "loss": 1.2441,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.3257335424423218,
      "learning_rate": 0.0015,
      "loss": 1.2429,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.349295049905777,
      "learning_rate": 0.0015,
      "loss": 1.2509,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.34802842140197754,
      "learning_rate": 0.0015,
      "loss": 1.2627,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.36794769763946533,
      "learning_rate": 0.0015,
      "loss": 1.2449,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.33518537878990173,
      "learning_rate": 0.0015,
      "loss": 1.2575,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.3931816518306732,
      "learning_rate": 0.0015,
      "loss": 1.2623,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.30917906761169434,
      "learning_rate": 0.0015,
      "loss": 1.2488,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.3136212229728699,
      "learning_rate": 0.0015,
      "loss": 1.2413,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.32720673084259033,
      "learning_rate": 0.0015,
      "loss": 1.2528,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.3788354694843292,
      "learning_rate": 0.0015,
      "loss": 1.2428,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.358934223651886,
      "learning_rate": 0.0015,
      "loss": 1.2455,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.387750506401062,
      "learning_rate": 0.0015,
      "loss": 1.2411,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.33726221323013306,
      "learning_rate": 0.0015,
      "loss": 1.2346,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 0.43859201669692993,
      "learning_rate": 0.0015,
      "loss": 1.2397,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.36601218581199646,
      "learning_rate": 0.0015,
      "loss": 1.2478,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.3264256715774536,
      "learning_rate": 0.0015,
      "loss": 1.2368,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.3220791220664978,
      "learning_rate": 0.0015,
      "loss": 1.248,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.35854560136795044,
      "learning_rate": 0.0015,
      "loss": 1.2464,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.3482915759086609,
      "learning_rate": 0.0015,
      "loss": 1.2305,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.39330121874809265,
      "learning_rate": 0.0015,
      "loss": 1.2477,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.36969903111457825,
      "learning_rate": 0.0015,
      "loss": 1.2547,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.31875529885292053,
      "learning_rate": 0.0015,
      "loss": 1.2408,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.33607614040374756,
      "learning_rate": 0.0015,
      "loss": 1.2444,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.3412076532840729,
      "learning_rate": 0.0015,
      "loss": 1.2279,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.35197487473487854,
      "learning_rate": 0.0015,
      "loss": 1.2474,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.3313758373260498,
      "learning_rate": 0.0015,
      "loss": 1.2353,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.38707274198532104,
      "learning_rate": 0.0015,
      "loss": 1.2499,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.36934447288513184,
      "learning_rate": 0.0015,
      "loss": 1.2423,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.33539143204689026,
      "learning_rate": 0.0015,
      "loss": 1.2413,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.3044580817222595,
      "learning_rate": 0.0015,
      "loss": 1.2364,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.34725913405418396,
      "learning_rate": 0.0015,
      "loss": 1.2338,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.32453039288520813,
      "learning_rate": 0.0015,
      "loss": 1.2468,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.31439530849456787,
      "learning_rate": 0.0015,
      "loss": 1.2395,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.3386889696121216,
      "learning_rate": 0.0015,
      "loss": 1.2394,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 0.39667123556137085,
      "learning_rate": 0.0015,
      "loss": 1.2369,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.35717153549194336,
      "learning_rate": 0.0015,
      "loss": 1.2481,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.4444461464881897,
      "learning_rate": 0.0015,
      "loss": 1.2463,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.4026486277580261,
      "learning_rate": 0.0014834368975312174,
      "loss": 1.2226,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.3683646321296692,
      "learning_rate": 0.0014629899726345957,
      "loss": 1.2439,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.3451366722583771,
      "learning_rate": 0.0014428248775471316,
      "loss": 1.2444,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.38098204135894775,
      "learning_rate": 0.00142293772767289,
      "loss": 1.2358,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.3486824631690979,
      "learning_rate": 0.001403324691959192,
      "loss": 1.2296,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.33603349328041077,
      "learning_rate": 0.0013839819921586025,
      "loss": 1.239,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.344035267829895,
      "learning_rate": 0.0013649059021010894,
      "loss": 1.2299,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.3314090073108673,
      "learning_rate": 0.0013460927469762154,
      "loss": 1.2303,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.3103720545768738,
      "learning_rate": 0.0013275389026252255,
      "loss": 1.2322,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.32174864411354065,
      "learning_rate": 0.0013092407948428887,
      "loss": 1.2229,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.3075444996356964,
      "learning_rate": 0.001291194898688966,
      "loss": 1.2325,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.32746440172195435,
      "learning_rate": 0.001273397737809166,
      "loss": 1.2281,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.3113197684288025,
      "learning_rate": 0.001255845883765463,
      "loss": 1.2253,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.3303135633468628,
      "learning_rate": 0.001238535955375642,
      "loss": 1.2178,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.30665114521980286,
      "learning_rate": 0.0012214646180619506,
      "loss": 1.2162,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.35169586539268494,
      "learning_rate": 0.001204628583208727,
      "loss": 1.2104,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.3337899446487427,
      "learning_rate": 0.0011880246075288827,
      "loss": 1.2201,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.30616530776023865,
      "learning_rate": 0.001171649492439115,
      "loss": 1.2112,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.3198193311691284,
      "learning_rate": 0.0011555000834437364,
      "loss": 1.2179,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.3075357675552368,
      "learning_rate": 0.0011395732695269908,
      "loss": 1.2106,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.3103102147579193,
      "learning_rate": 0.0011238659825537505,
      "loss": 1.1934,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.35869720578193665,
      "learning_rate": 0.0011083751966784717,
      "loss": 1.1977,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.33826592564582825,
      "learning_rate": 0.0010930979277622953,
      "loss": 1.2129,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.31560054421424866,
      "learning_rate": 0.0010780312327981854,
      "loss": 1.2108,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.31668636202812195,
      "learning_rate": 0.0010631722093439888,
      "loss": 1.2027,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.32392340898513794,
      "learning_rate": 0.00104851799496331,
      "loss": 1.1943,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.35339978337287903,
      "learning_rate": 0.0010340657666740914,
      "loss": 1.2038,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.35348406434059143,
      "learning_rate": 0.0010198127404047975,
      "loss": 1.184,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.33255070447921753,
      "learning_rate": 0.0010057561704580897,
      "loss": 1.1903,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.3151698410511017,
      "learning_rate": 0.0009918933489818985,
      "loss": 1.1994,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.3220020830631256,
      "learning_rate": 0.0009782216054477827,
      "loss": 1.1953,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.33605191111564636,
      "learning_rate": 0.0009647383061364801,
      "loss": 1.1995,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.3029930889606476,
      "learning_rate": 0.0009514408536305495,
      "loss": 1.1926,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.29145169258117676,
      "learning_rate": 0.0009383266863140042,
      "loss": 1.2068,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.35904476046562195,
      "learning_rate": 0.000925393277878844,
      "loss": 1.2043,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.3455471098423004,
      "learning_rate": 0.0009126381368383879,
      "loss": 1.1899,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.3116503357887268,
      "learning_rate": 0.0009000588060473156,
      "loss": 1.1791,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.31568071246147156,
      "learning_rate": 0.0008876528622283235,
      "loss": 1.1899,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.35425931215286255,
      "learning_rate": 0.0008754179155053053,
      "loss": 1.1867,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.3980303704738617,
      "learning_rate": 0.0008633516089429683,
      "loss": 1.1843,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.36593008041381836,
      "learning_rate": 0.0008514516180927928,
      "loss": 1.1812,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.32517319917678833,
      "learning_rate": 0.0008397156505452524,
      "loss": 1.1772,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.325745552778244,
      "learning_rate": 0.0008281414454882051,
      "loss": 1.1787,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.2857639193534851,
      "learning_rate": 0.0008167267732713704,
      "loss": 1.1855,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.3474614918231964,
      "learning_rate": 0.0008054694349768117,
      "loss": 1.1699,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.3178200125694275,
      "learning_rate": 0.0007943672619953359,
      "loss": 1.1795,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.3442220389842987,
      "learning_rate": 0.0007834181156087356,
      "loss": 1.1726,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.32902151346206665,
      "learning_rate": 0.0007726198865777852,
      "loss": 1.1736,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.3253075182437897,
      "learning_rate": 0.0007619704947359191,
      "loss": 1.1703,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.31004899740219116,
      "learning_rate": 0.0007514678885885087,
      "loss": 1.1713,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.32050368189811707,
      "learning_rate": 0.0007411100449176633,
      "loss": 1.1705,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.296805202960968,
      "learning_rate": 0.0007308949683924791,
      "loss": 1.1722,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.3058995008468628,
      "learning_rate": 0.000720820691184658,
      "loss": 1.1659,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.31293025612831116,
      "learning_rate": 0.0007108852725894269,
      "loss": 1.1655,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.2998400926589966,
      "learning_rate": 0.000701086798651681,
      "loss": 1.1686,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.30473223328590393,
      "learning_rate": 0.0006914233817972798,
      "loss": 1.158,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.32161760330200195,
      "learning_rate": 0.0006818931604694261,
      "loss": 1.1622,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.30090296268463135,
      "learning_rate": 0.0006724942987700563,
      "loss": 1.1657,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.32431352138519287,
      "learning_rate": 0.0006632249861061732,
      "loss": 1.1663,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.3277837038040161,
      "learning_rate": 0.0006540834368410549,
      "loss": 1.1631,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.3163105249404907,
      "learning_rate": 0.0006450678899502701,
      "loss": 1.1637,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.3263886868953705,
      "learning_rate": 0.0006361766086824345,
      "loss": 1.1617,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.3312484622001648,
      "learning_rate": 0.000627407880224645,
      "loss": 1.1657,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.30463480949401855,
      "learning_rate": 0.0006187600153725225,
      "loss": 1.1532,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.3175445795059204,
      "learning_rate": 0.0006102313482048055,
      "loss": 1.1563,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.32396236062049866,
      "learning_rate": 0.0006018202357624274,
      "loss": 1.1583,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.29472267627716064,
      "learning_rate": 0.0005935250577320168,
      "loss": 1.1567,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.33541539311408997,
      "learning_rate": 0.0005853442161337618,
      "loss": 1.144,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.29096779227256775,
      "learning_rate": 0.0005772761350135759,
      "loss": 1.154,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.31761106848716736,
      "learning_rate": 0.0005693192601395058,
      "loss": 1.1484,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.3081628084182739,
      "learning_rate": 0.000561472058702326,
      "loss": 1.1436,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.30715107917785645,
      "learning_rate": 0.000553733019020258,
      "loss": 1.1543,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.30482837557792664,
      "learning_rate": 0.0005461006502477612,
      "loss": 1.1422,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.3131371736526489,
      "learning_rate": 0.0005385734820883369,
      "loss": 1.1417,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.30512696504592896,
      "learning_rate": 0.0005311500645112907,
      "loss": 1.1661,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.3093601167201996,
      "learning_rate": 0.0005238289674723993,
      "loss": 1.1446,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.3076038658618927,
      "learning_rate": 0.0005166087806384274,
      "loss": 1.1486,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.305584192276001,
      "learning_rate": 0.0005094881131154418,
      "loss": 1.1522,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.31308478116989136,
      "learning_rate": 0.0005024655931808696,
      "loss": 1.1483,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.3094158470630646,
      "learning_rate": 0.0004955398680192508,
      "loss": 1.1419,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.29979753494262695,
      "learning_rate": 0.000488709603461632,
      "loss": 1.1373,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.2946663200855255,
      "learning_rate": 0.000481973483728553,
      "loss": 1.1424,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.30663517117500305,
      "learning_rate": 0.0004753302111765748,
      "loss": 1.14,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.3226993680000305,
      "learning_rate": 0.0004687785060483032,
      "loss": 1.1494,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.2898540198802948,
      "learning_rate": 0.0004623171062258558,
      "loss": 1.123,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.3063409626483917,
      "learning_rate": 0.0004559447669877288,
      "loss": 1.1363,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.2905513346195221,
      "learning_rate": 0.00044966026076901413,
      "loss": 1.1411,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.34559938311576843,
      "learning_rate": 0.00044346237692492177,
      "loss": 1.1461,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.3467509150505066,
      "learning_rate": 0.0004373499214975615,
      "loss": 1.1328,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.3333624601364136,
      "learning_rate": 0.0004313217169859396,
      "loss": 1.1366,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.3122444748878479,
      "learning_rate": 0.0004253766021191256,
      "loss": 1.1423,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.3039610683917999,
      "learning_rate": 0.00041951343163254497,
      "loss": 1.1379,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.301398903131485,
      "learning_rate": 0.00041373107604735626,
      "loss": 1.1351,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.3017662465572357,
      "learning_rate": 0.0004080284214528687,
      "loss": 1.1331,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3178175687789917,
      "learning_rate": 0.0004024043692919589,
      "loss": 1.1421,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.34835416078567505,
      "learning_rate": 0.0003968578361494449,
      "loss": 1.1341,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.32286974787712097,
      "learning_rate": 0.000391387753543378,
      "loss": 1.1443,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.3165818750858307,
      "learning_rate": 0.00038599306771921023,
      "loss": 1.1267,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.32807934284210205,
      "learning_rate": 0.0003806727394468004,
      "loss": 1.1253,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.30154094099998474,
      "learning_rate": 0.0003754257438202162,
      "loss": 1.1331,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.3244779407978058,
      "learning_rate": 0.0003702510700602974,
      "loss": 1.1407,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.35364311933517456,
      "learning_rate": 0.0003651477213199393,
      "loss": 1.1187,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.31307119131088257,
      "learning_rate": 0.000360114714492061,
      "loss": 1.1218,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.3015627861022949,
      "learning_rate": 0.0003551510800202195,
      "loss": 1.1251,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.37355753779411316,
      "learning_rate": 0.0003502558617118353,
      "loss": 1.1302,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.30779457092285156,
      "learning_rate": 0.0003454281165539914,
      "loss": 1.1422,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.31253111362457275,
      "learning_rate": 0.00034066691453177176,
      "loss": 1.139,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.29553452134132385,
      "learning_rate": 0.0003359713384491037,
      "loss": 1.1353,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.3091430366039276,
      "learning_rate": 0.00033134048375206944,
      "loss": 1.1296,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.35152044892311096,
      "learning_rate": 0.0003267734583546536,
      "loss": 1.1285,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.30567291378974915,
      "learning_rate": 0.00032226938246689157,
      "loss": 1.1239,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.3052709400653839,
      "learning_rate": 0.0003178273884253874,
      "loss": 1.1263,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.3217529058456421,
      "learning_rate": 0.0003134466205261674,
      "loss": 1.1366,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.29919829964637756,
      "learning_rate": 0.0003091262348598378,
      "loss": 1.1403,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.29513537883758545,
      "learning_rate": 0.0003048653991490141,
      "loss": 1.1203,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.3004360795021057,
      "learning_rate": 0.00030066329258799187,
      "loss": 1.1179,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.3202657401561737,
      "learning_rate": 0.0002965191056846266,
      "loss": 1.1212,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.3142313063144684,
      "learning_rate": 0.000292432040104394,
      "loss": 1.1143,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.3124704957008362,
      "learning_rate": 0.00028840130851659853,
      "loss": 1.1158,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.3039351999759674,
      "learning_rate": 0.0002844261344427028,
      "loss": 1.1249,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.3144666254520416,
      "learning_rate": 0.0002805057521067471,
      "loss": 1.1244,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.29907071590423584,
      "learning_rate": 0.00027663940628783017,
      "loss": 1.1107,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.30688223242759705,
      "learning_rate": 0.00027282635217462393,
      "loss": 1.1208,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.31314346194267273,
      "learning_rate": 0.0002690658552218937,
      "loss": 1.1282,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.2993510663509369,
      "learning_rate": 0.00026535719100899516,
      "loss": 1.1084,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.31194618344306946,
      "learning_rate": 0.00026169964510032245,
      "loss": 1.1114,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.2804999053478241,
      "learning_rate": 0.00025809251290767984,
      "loss": 1.1028,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.29775840044021606,
      "learning_rate": 0.00025453509955454957,
      "loss": 1.1051,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.3113470673561096,
      "learning_rate": 0.00025102671974223175,
      "loss": 1.1086,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.30034735798835754,
      "learning_rate": 0.00024756669761782815,
      "loss": 1.1175,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.31758636236190796,
      "learning_rate": 0.0002441543666440464,
      "loss": 1.1079,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.29403379559516907,
      "learning_rate": 0.00024078906947079878,
      "loss": 1.1159,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.30131855607032776,
      "learning_rate": 0.00023747015780857005,
      "loss": 1.1213,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.3080956041812897,
      "learning_rate": 0.00023419699230353144,
      "loss": 1.1184,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.3078226149082184,
      "learning_rate": 0.00023096894241437586,
      "loss": 1.1216,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.2988618016242981,
      "learning_rate": 0.00022778538629085056,
      "loss": 1.1068,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.2947309613227844,
      "learning_rate": 0.00022464571065396427,
      "loss": 1.1092,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.330911785364151,
      "learning_rate": 0.00022154931067784521,
      "loss": 1.104,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.3285962641239166,
      "learning_rate": 0.00021849558987322782,
      "loss": 1.1035,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.3116293251514435,
      "learning_rate": 0.0002154839599725452,
      "loss": 1.1075,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.3200410008430481,
      "learning_rate": 0.00021251384081660544,
      "loss": 1.1139,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.2930144667625427,
      "learning_rate": 0.0002095846602428303,
      "loss": 1.1111,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.29921388626098633,
      "learning_rate": 0.00020669585397503358,
      "loss": 1.1022,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.2902184724807739,
      "learning_rate": 0.0002038468655147195,
      "loss": 1.1076,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.2966500520706177,
      "learning_rate": 0.00020103714603387894,
      "loss": 1.1203,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.28531065583229065,
      "learning_rate": 0.00019826615426926338,
      "loss": 1.097,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.2880052328109741,
      "learning_rate": 0.00019553335641811625,
      "loss": 1.1186,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.3241575360298157,
      "learning_rate": 0.0001928382260353415,
      "loss": 1.1088,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.29371729493141174,
      "learning_rate": 0.00019018024393208902,
      "loss": 1.1235,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.2925833761692047,
      "learning_rate": 0.00018755889807573872,
      "loss": 1.101,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.29287418723106384,
      "learning_rate": 0.00018497368349126262,
      "loss": 1.1088,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.3102490305900574,
      "learning_rate": 0.00018242410216394648,
      "loss": 1.1191,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.3129653334617615,
      "learning_rate": 0.0001799096629434529,
      "loss": 1.0939,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.3090495467185974,
      "learning_rate": 0.00017742988144920578,
      "loss": 1.1038,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.2979944050312042,
      "learning_rate": 0.00017498427997707976,
      "loss": 1.1031,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.2943827211856842,
      "learning_rate": 0.00017257238740737548,
      "loss": 1.1057,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.3144172132015228,
      "learning_rate": 0.00017019373911406307,
      "loss": 1.1136,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.30515575408935547,
      "learning_rate": 0.000167847876875277,
      "loss": 1.1112,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.30309703946113586,
      "learning_rate": 0.00016553434878504428,
      "loss": 1.0985,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.3171931505203247,
      "learning_rate": 0.00016325270916622947,
      "loss": 1.0964,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.30823394656181335,
      "learning_rate": 0.00016100251848467966,
      "loss": 1.1039,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.30423763394355774,
      "learning_rate": 0.0001587833432645528,
      "loss": 1.0993,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.31251442432403564,
      "learning_rate": 0.00015659475600481292,
      "loss": 1.1133,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.28880876302719116,
      "learning_rate": 0.00015443633509687688,
      "loss": 1.1037,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.2903044819831848,
      "learning_rate": 0.00015230766474339536,
      "loss": 1.1018,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.29660487174987793,
      "learning_rate": 0.00015020833487815416,
      "loss": 1.1117,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.30001166462898254,
      "learning_rate": 0.0001481379410870792,
      "loss": 1.1012,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.29828891158103943,
      "learning_rate": 0.00014609608453033013,
      "loss": 1.0863,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.29176872968673706,
      "learning_rate": 0.00014408237186546807,
      "loss": 1.0958,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.2964164614677429,
      "learning_rate": 0.00014209641517168273,
      "loss": 1.0823,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.33330321311950684,
      "learning_rate": 0.00014013783187506265,
      "loss": 1.0987,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.30649515986442566,
      "learning_rate": 0.00013820624467489697,
      "loss": 1.1167,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.29155051708221436,
      "learning_rate": 0.00013630128147099213,
      "loss": 1.1118,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.3033908009529114,
      "learning_rate": 0.00013442257529199068,
      "loss": 1.0933,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.30457115173339844,
      "learning_rate": 0.00013256976422467803,
      "loss": 1.1022,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.29595619440078735,
      "learning_rate": 0.00013074249134426366,
      "loss": 1.0985,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.3050054609775543,
      "learning_rate": 0.0001289404046456233,
      "loss": 1.1071,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.2907356023788452,
      "learning_rate": 0.0001271631569754887,
      "loss": 1.1019,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.32138654589653015,
      "learning_rate": 0.0001254104059655723,
      "loss": 1.1048,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.29230400919914246,
      "learning_rate": 0.00012368181396661337,
      "loss": 1.0924,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.3035050928592682,
      "learning_rate": 0.00012197704798333364,
      "loss": 1.0923,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.294167160987854,
      "learning_rate": 0.00012029577961028894,
      "loss": 1.092,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.3287489414215088,
      "learning_rate": 0.00011863768496860542,
      "loss": 1.1029,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.2865371108055115,
      "learning_rate": 0.00011700244464358777,
      "loss": 1.0994,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.293733149766922,
      "learning_rate": 0.00011538974362318715,
      "loss": 1.102,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.2847878038883209,
      "learning_rate": 0.00011379927123731737,
      "loss": 1.0974,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.29861894249916077,
      "learning_rate": 0.0001122307210980077,
      "loss": 1.0946,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.28597918152809143,
      "learning_rate": 0.00011068379104038026,
      "loss": 1.1079,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.3035583198070526,
      "learning_rate": 0.00010915818306444116,
      "loss": 1.0885,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.2900126278400421,
      "learning_rate": 0.00010765360327767384,
      "loss": 1.0977,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.2991573214530945,
      "learning_rate": 0.00010616976183842376,
      "loss": 1.1008,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.29756081104278564,
      "learning_rate": 0.00010470637290006365,
      "loss": 1.1022,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.2888975441455841,
      "learning_rate": 0.00010326315455592764,
      "loss": 1.093,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.28867655992507935,
      "learning_rate": 0.0001018398287850053,
      "loss": 1.089,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.3160809278488159,
      "learning_rate": 0.00010043612139838357,
      "loss": 1.1028,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.29105138778686523,
      "learning_rate": 9.905176198642719e-05,
      "loss": 1.0934,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.3039015829563141,
      "learning_rate": 9.76864838666871e-05,
      "loss": 1.095,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.3249395489692688,
      "learning_rate": 9.634002403252676e-05,
      "loss": 1.0907,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.30835098028182983,
      "learning_rate": 9.501212310245681e-05,
      "loss": 1.0901,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.3204496502876282,
      "learning_rate": 9.370252527016777e-05,
      "loss": 1.0998,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.29987505078315735,
      "learning_rate": 9.241097825525163e-05,
      "loss": 1.0899,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.30191531777381897,
      "learning_rate": 9.113723325460276e-05,
      "loss": 1.0984,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.305660218000412,
      "learning_rate": 8.988104489448849e-05,
      "loss": 1.0915,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.30820146203041077,
      "learning_rate": 8.864217118328042e-05,
      "loss": 1.1043,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.3056180775165558,
      "learning_rate": 8.742037346483729e-05,
      "loss": 1.0953,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.3009982109069824,
      "learning_rate": 8.62154163725303e-05,
      "loss": 1.1028,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.2943590581417084,
      "learning_rate": 8.502706778390219e-05,
      "loss": 1.098,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.2908317446708679,
      "learning_rate": 8.38550987759513e-05,
      "loss": 1.0953,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.2977442145347595,
      "learning_rate": 8.269928358103191e-05,
      "loss": 1.1083,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.30347803235054016,
      "learning_rate": 8.155939954336243e-05,
      "loss": 1.0993,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.2959662675857544,
      "learning_rate": 8.043522707613312e-05,
      "loss": 1.0974,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.2978464365005493,
      "learning_rate": 7.932654961920486e-05,
      "loss": 1.081,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.30118417739868164,
      "learning_rate": 7.823315359739135e-05,
      "loss": 1.0832,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.29306381940841675,
      "learning_rate": 7.715482837931577e-05,
      "loss": 1.1044,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.30623766779899597,
      "learning_rate": 7.6091366236835e-05,
      "loss": 1.0839,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.2869498133659363,
      "learning_rate": 7.504256230502289e-05,
      "loss": 1.1019,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.2981465756893158,
      "learning_rate": 7.400821454270524e-05,
      "loss": 1.0953,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.29135996103286743,
      "learning_rate": 7.29881236935386e-05,
      "loss": 1.0858,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.2871904969215393,
      "learning_rate": 7.198209324762562e-05,
      "loss": 1.0872,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.2934054732322693,
      "learning_rate": 7.098992940365946e-05,
      "loss": 1.0869,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.2883404791355133,
      "learning_rate": 7.001144103159e-05,
      "loss": 1.0924,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.2962549924850464,
      "learning_rate": 6.904643963580461e-05,
      "loss": 1.0977,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.3258002996444702,
      "learning_rate": 6.809473931881644e-05,
      "loss": 1.0899,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.2951005697250366,
      "learning_rate": 6.71561567454532e-05,
      "loss": 1.0893,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.29630401730537415,
      "learning_rate": 6.623051110753948e-05,
      "loss": 1.106,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.2885693311691284,
      "learning_rate": 6.531762408906607e-05,
      "loss": 1.096,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.2941160500049591,
      "learning_rate": 6.441731983183912e-05,
      "loss": 1.0941,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.30166760087013245,
      "learning_rate": 6.352942490160292e-05,
      "loss": 1.0857,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.3127710521221161,
      "learning_rate": 6.265376825462966e-05,
      "loss": 1.0941,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.29389771819114685,
      "learning_rate": 6.179018120476945e-05,
      "loss": 1.0892,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.29795289039611816,
      "learning_rate": 6.0938497390954946e-05,
      "loss": 1.091,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.2943289577960968,
      "learning_rate": 6.009855274515339e-05,
      "loss": 1.0861,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.2887933850288391,
      "learning_rate": 5.9270185460760735e-05,
      "loss": 1.0902,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.2993631660938263,
      "learning_rate": 5.8453235961431225e-05,
      "loss": 1.0819,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.28679993748664856,
      "learning_rate": 5.764754687033678e-05,
      "loss": 1.0915,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.29990583658218384,
      "learning_rate": 5.6852962979849836e-05,
      "loss": 1.0873,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.29532426595687866,
      "learning_rate": 5.6069331221644284e-05,
      "loss": 1.1018,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.2920057475566864,
      "learning_rate": 5.529650063720842e-05,
      "loss": 1.0974,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.30257919430732727,
      "learning_rate": 5.453432234876445e-05,
      "loss": 1.0905,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.2978675663471222,
      "learning_rate": 5.37826495305886e-05,
      "loss": 1.0806,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.3190942406654358,
      "learning_rate": 5.304133738072674e-05,
      "loss": 1.0998,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.2860528826713562,
      "learning_rate": 5.2310243093099814e-05,
      "loss": 1.0911,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.3035094738006592,
      "learning_rate": 5.158922582999368e-05,
      "loss": 1.094,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.2932252287864685,
      "learning_rate": 5.087814669492819e-05,
      "loss": 1.0858,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.3255833685398102,
      "learning_rate": 5.017686870590028e-05,
      "loss": 1.0896,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.2977413237094879,
      "learning_rate": 4.948525676899577e-05,
      "loss": 1.0875,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.31040334701538086,
      "learning_rate": 4.880317765236493e-05,
      "loss": 1.0948,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.31936654448509216,
      "learning_rate": 4.8130499960556755e-05,
      "loss": 1.0764,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.3022034764289856,
      "learning_rate": 4.746709410920699e-05,
      "loss": 1.0828,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.29438066482543945,
      "learning_rate": 4.681283230007507e-05,
      "loss": 1.0821,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.2978823184967041,
      "learning_rate": 4.616758849642509e-05,
      "loss": 1.094,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.28286972641944885,
      "learning_rate": 4.553123839874615e-05,
      "loss": 1.0975,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.2875301241874695,
      "learning_rate": 4.490365942080736e-05,
      "loss": 1.0921,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.28754597902297974,
      "learning_rate": 4.428473066604285e-05,
      "loss": 1.0898,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.3019983172416687,
      "learning_rate": 4.367433290426233e-05,
      "loss": 1.0886,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.3068964183330536,
      "learning_rate": 4.3072348548682595e-05,
      "loss": 1.0845,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.3060677647590637,
      "learning_rate": 4.247866163327575e-05,
      "loss": 1.0976,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.30172476172447205,
      "learning_rate": 4.1893157790429404e-05,
      "loss": 1.0873,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.30481332540512085,
      "learning_rate": 4.1315724228915066e-05,
      "loss": 1.0837,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.2977677583694458,
      "learning_rate": 4.074624971216005e-05,
      "loss": 1.0734,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.28608590364456177,
      "learning_rate": 4.018462453681889e-05,
      "loss": 1.0808,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.2980504035949707,
      "learning_rate": 3.963074051164014e-05,
      "loss": 1.0892,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.30341869592666626,
      "learning_rate": 3.908449093662446e-05,
      "loss": 1.0849,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.2970033884048462,
      "learning_rate": 3.854577058246998e-05,
      "loss": 1.0791,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.2947596609592438,
      "learning_rate": 3.801447567030094e-05,
      "loss": 1.0989,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.28587204217910767,
      "learning_rate": 3.7490503851675777e-05,
      "loss": 1.0862,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.30299025774002075,
      "learning_rate": 3.6973754188870806e-05,
      "loss": 1.1024,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.28645822405815125,
      "learning_rate": 3.6464127135435536e-05,
      "loss": 1.0886,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.3179738223552704,
      "learning_rate": 3.596152451701616e-05,
      "loss": 1.0871,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.3064330518245697,
      "learning_rate": 3.5465849512443226e-05,
      "loss": 1.0874,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.2956417500972748,
      "learning_rate": 3.4977006635080086e-05,
      "loss": 1.0883,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.29473647475242615,
      "learning_rate": 3.449490171442838e-05,
      "loss": 1.0936,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.29144287109375,
      "learning_rate": 3.401944187798702e-05,
      "loss": 1.091,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.2938981354236603,
      "learning_rate": 3.355053553336137e-05,
      "loss": 1.0807,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.3047492504119873,
      "learning_rate": 3.308809235061882e-05,
      "loss": 1.0822,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.29885032773017883,
      "learning_rate": 3.263202324488772e-05,
      "loss": 1.0891,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.28590863943099976,
      "learning_rate": 3.218224035919609e-05,
      "loss": 1.0853,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.29893019795417786,
      "learning_rate": 3.173865704754688e-05,
      "loss": 1.084,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.28867775201797485,
      "learning_rate": 3.130118785822657e-05,
      "loss": 1.0881,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.30242010951042175,
      "learning_rate": 3.08697485173437e-05,
      "loss": 1.0897,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.30233702063560486,
      "learning_rate": 3.0444255912594442e-05,
      "loss": 1.0869,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.30981701612472534,
      "learning_rate": 3.002462807725185e-05,
      "loss": 1.0843,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.3049086630344391,
      "learning_rate": 2.9610784174375868e-05,
      "loss": 1.0892,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.2946826219558716,
      "learning_rate": 2.920264448124087e-05,
      "loss": 1.0872,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.30069807171821594,
      "learning_rate": 2.8800130373977936e-05,
      "loss": 1.081,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.28993991017341614,
      "learning_rate": 2.84031643124288e-05,
      "loss": 1.0833,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.3197329342365265,
      "learning_rate": 2.8011669825208517e-05,
      "loss": 1.1017,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.30109071731567383,
      "learning_rate": 2.762557149497405e-05,
      "loss": 1.0805,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.2971667945384979,
      "learning_rate": 2.724479494389592e-05,
      "loss": 1.0882,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.2833009362220764,
      "learning_rate": 2.6869266819330058e-05,
      "loss": 1.0903,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.2999628186225891,
      "learning_rate": 2.6498914779687228e-05,
      "loss": 1.0894,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.29329007863998413,
      "learning_rate": 2.6133667480497115e-05,
      "loss": 1.0934,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.3185789883136749,
      "learning_rate": 2.5773454560664597e-05,
      "loss": 1.0925,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.3005650043487549,
      "learning_rate": 2.541820662891541e-05,
      "loss": 1.0723,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.2987126410007477,
      "learning_rate": 2.5067855250428616e-05,
      "loss": 1.082,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.28658491373062134,
      "learning_rate": 2.472233293365335e-05,
      "loss": 1.0943,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.2930833697319031,
      "learning_rate": 2.4381573117307307e-05,
      "loss": 1.0793,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.31947436928749084,
      "learning_rate": 2.4045510157554362e-05,
      "loss": 1.0969,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.3002609312534332,
      "learning_rate": 2.3714079315358985e-05,
      "loss": 1.0834,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.29156044125556946,
      "learning_rate": 2.338721674401494e-05,
      "loss": 1.091,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.30455270409584045,
      "learning_rate": 2.30648594768459e-05,
      "loss": 1.0899,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.2864668667316437,
      "learning_rate": 2.2746945415075523e-05,
      "loss": 1.0872,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.3029510974884033,
      "learning_rate": 2.2433413315864803e-05,
      "loss": 1.083,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.2998829483985901,
      "learning_rate": 2.2124202780514277e-05,
      "loss": 1.0851,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.2927697002887726,
      "learning_rate": 2.1819254242828815e-05,
      "loss": 1.0858,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.29847201704978943,
      "learning_rate": 2.151850895764285e-05,
      "loss": 1.0816,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.291204035282135,
      "learning_rate": 2.12219089895037e-05,
      "loss": 1.0765,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.3079606592655182,
      "learning_rate": 2.092939720151092e-05,
      "loss": 1.0721,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.3021531403064728,
      "learning_rate": 2.064091724430947e-05,
      "loss": 1.0819,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.2918916642665863,
      "learning_rate": 2.0356413545234603e-05,
      "loss": 1.0785,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.30379199981689453,
      "learning_rate": 2.0075831297606357e-05,
      "loss": 1.0873,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.29026755690574646,
      "learning_rate": 1.9799116450171627e-05,
      "loss": 1.0746,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.28747591376304626,
      "learning_rate": 1.952621569669175e-05,
      "loss": 1.0864,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.2957964539527893,
      "learning_rate": 1.9257076465673605e-05,
      "loss": 1.0824,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.2994972765445709,
      "learning_rate": 1.899164691024229e-05,
      "loss": 1.0846,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.30032384395599365,
      "learning_rate": 1.872987589815331e-05,
      "loss": 1.0809,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.31126585602760315,
      "learning_rate": 1.8471713001942538e-05,
      "loss": 1.0893,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.30770644545555115,
      "learning_rate": 1.8217108489211845e-05,
      "loss": 1.0888,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.2887580990791321,
      "learning_rate": 1.7966013313048696e-05,
      "loss": 1.0989,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.299443781375885,
      "learning_rate": 1.7718379102577752e-05,
      "loss": 1.0882,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.29414528608322144,
      "learning_rate": 1.7474158153642745e-05,
      "loss": 1.087,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.2917827069759369,
      "learning_rate": 1.7233303419616745e-05,
      "loss": 1.0745,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.3026796877384186,
      "learning_rate": 1.699576850233916e-05,
      "loss": 1.0796,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.3261394798755646,
      "learning_rate": 1.6761507643177553e-05,
      "loss": 1.094,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.29815807938575745,
      "learning_rate": 1.6530475714212752e-05,
      "loss": 1.0693,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.29984185099601746,
      "learning_rate": 1.6302628209545423e-05,
      "loss": 1.0877,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.30411556363105774,
      "learning_rate": 1.6077921236722464e-05,
      "loss": 1.0886,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.28755223751068115,
      "learning_rate": 1.5856311508281594e-05,
      "loss": 1.0829,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.2939528226852417,
      "learning_rate": 1.5637756333412454e-05,
      "loss": 1.0933,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.2983037233352661,
      "learning_rate": 1.542221360973268e-05,
      "loss": 1.0775,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.29906195402145386,
      "learning_rate": 1.5209641815177312e-05,
      "loss": 1.09,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.8255208134651184,
      "learning_rate": 1.5e-05,
      "loss": 1.0781,
      "step": 9480
    }
  ],
  "logging_steps": 10,
  "max_steps": 9480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.036240179760947e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}