{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.17250301880282906,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0003450060376056581,
      "grad_norm": 34.52312469482422,
      "learning_rate": 2.02e-06,
      "loss": 15.0889,
      "step": 1
    },
    {
      "epoch": 0.0006900120752113162,
      "grad_norm": 36.38729476928711,
      "learning_rate": 4.04e-06,
      "loss": 15.1439,
      "step": 2
    },
    {
      "epoch": 0.0010350181128169743,
      "grad_norm": 38.78995895385742,
      "learning_rate": 6.06e-06,
      "loss": 15.8469,
      "step": 3
    },
    {
      "epoch": 0.0013800241504226323,
      "grad_norm": 43.361572265625,
      "learning_rate": 8.08e-06,
      "loss": 15.6281,
      "step": 4
    },
    {
      "epoch": 0.0017250301880282904,
      "grad_norm": 40.78726577758789,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 16.8042,
      "step": 5
    },
    {
      "epoch": 0.0020700362256339485,
      "grad_norm": 44.38218307495117,
      "learning_rate": 1.212e-05,
      "loss": 16.4606,
      "step": 6
    },
    {
      "epoch": 0.002415042263239607,
      "grad_norm": 40.80599594116211,
      "learning_rate": 1.4140000000000002e-05,
      "loss": 15.4033,
      "step": 7
    },
    {
      "epoch": 0.0027600483008452647,
      "grad_norm": 44.74884796142578,
      "learning_rate": 1.616e-05,
      "loss": 14.2032,
      "step": 8
    },
    {
      "epoch": 0.003105054338450923,
      "grad_norm": 42.170570373535156,
      "learning_rate": 1.818e-05,
      "loss": 13.8714,
      "step": 9
    },
    {
      "epoch": 0.003450060376056581,
      "grad_norm": 37.16923904418945,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 12.1789,
      "step": 10
    },
    {
      "epoch": 0.003795066413662239,
      "grad_norm": 36.612850189208984,
      "learning_rate": 2.222e-05,
      "loss": 11.7298,
      "step": 11
    },
    {
      "epoch": 0.004140072451267897,
      "grad_norm": 29.61734390258789,
      "learning_rate": 2.424e-05,
      "loss": 12.3502,
      "step": 12
    },
    {
      "epoch": 0.004485078488873555,
      "grad_norm": 28.27483367919922,
      "learning_rate": 2.6260000000000003e-05,
      "loss": 11.2876,
      "step": 13
    },
    {
      "epoch": 0.004830084526479214,
      "grad_norm": 25.60624122619629,
      "learning_rate": 2.8280000000000004e-05,
      "loss": 11.1896,
      "step": 14
    },
    {
      "epoch": 0.005175090564084871,
      "grad_norm": 25.410795211791992,
      "learning_rate": 3.0299999999999998e-05,
      "loss": 9.7695,
      "step": 15
    },
    {
      "epoch": 0.005520096601690529,
      "grad_norm": 23.130945205688477,
      "learning_rate": 3.232e-05,
      "loss": 10.6948,
      "step": 16
    },
    {
      "epoch": 0.005865102639296188,
      "grad_norm": 26.821157455444336,
      "learning_rate": 3.434e-05,
      "loss": 10.7685,
      "step": 17
    },
    {
      "epoch": 0.006210108676901846,
      "grad_norm": 25.83894157409668,
      "learning_rate": 3.636e-05,
      "loss": 10.3437,
      "step": 18
    },
    {
      "epoch": 0.006555114714507504,
      "grad_norm": 22.30074691772461,
      "learning_rate": 3.838e-05,
      "loss": 9.5938,
      "step": 19
    },
    {
      "epoch": 0.006900120752113162,
      "grad_norm": 22.2345027923584,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 10.5932,
      "step": 20
    },
    {
      "epoch": 0.00724512678971882,
      "grad_norm": 23.70596694946289,
      "learning_rate": 4.242e-05,
      "loss": 10.0185,
      "step": 21
    },
    {
      "epoch": 0.007590132827324478,
      "grad_norm": 19.10881233215332,
      "learning_rate": 4.444e-05,
      "loss": 9.4553,
      "step": 22
    },
    {
      "epoch": 0.007935138864930136,
      "grad_norm": 19.092126846313477,
      "learning_rate": 4.6460000000000006e-05,
      "loss": 9.2582,
      "step": 23
    },
    {
      "epoch": 0.008280144902535794,
      "grad_norm": 17.3260555267334,
      "learning_rate": 4.848e-05,
      "loss": 9.7332,
      "step": 24
    },
    {
      "epoch": 0.008625150940141452,
      "grad_norm": 16.65959930419922,
      "learning_rate": 5.05e-05,
      "loss": 9.1777,
      "step": 25
    },
    {
      "epoch": 0.00897015697774711,
      "grad_norm": 17.30274772644043,
      "learning_rate": 5.2520000000000005e-05,
      "loss": 9.842,
      "step": 26
    },
    {
      "epoch": 0.009315163015352769,
      "grad_norm": 19.711753845214844,
      "learning_rate": 5.454e-05,
      "loss": 9.3012,
      "step": 27
    },
    {
      "epoch": 0.009660169052958427,
      "grad_norm": 18.88648796081543,
      "learning_rate": 5.656000000000001e-05,
      "loss": 9.838,
      "step": 28
    },
    {
      "epoch": 0.010005175090564086,
      "grad_norm": 19.750307083129883,
      "learning_rate": 5.858e-05,
      "loss": 10.2961,
      "step": 29
    },
    {
      "epoch": 0.010350181128169742,
      "grad_norm": 18.343475341796875,
      "learning_rate": 6.0599999999999996e-05,
      "loss": 8.6186,
      "step": 30
    },
    {
      "epoch": 0.0106951871657754,
      "grad_norm": 15.626960754394531,
      "learning_rate": 6.262000000000001e-05,
      "loss": 9.9093,
      "step": 31
    },
    {
      "epoch": 0.011040193203381059,
      "grad_norm": 16.598981857299805,
      "learning_rate": 6.464e-05,
      "loss": 9.3308,
      "step": 32
    },
    {
      "epoch": 0.011385199240986717,
      "grad_norm": 17.307716369628906,
      "learning_rate": 6.666e-05,
      "loss": 9.4876,
      "step": 33
    },
    {
      "epoch": 0.011730205278592375,
      "grad_norm": 15.73810863494873,
      "learning_rate": 6.868e-05,
      "loss": 9.8741,
      "step": 34
    },
    {
      "epoch": 0.012075211316198034,
      "grad_norm": 16.63450050354004,
      "learning_rate": 7.07e-05,
      "loss": 9.2182,
      "step": 35
    },
    {
      "epoch": 0.012420217353803692,
      "grad_norm": 15.06506061553955,
      "learning_rate": 7.272e-05,
      "loss": 9.2064,
      "step": 36
    },
    {
      "epoch": 0.01276522339140935,
      "grad_norm": 16.938413619995117,
      "learning_rate": 7.474e-05,
      "loss": 9.5759,
      "step": 37
    },
    {
      "epoch": 0.013110229429015009,
      "grad_norm": 15.28792953491211,
      "learning_rate": 7.676e-05,
      "loss": 9.4626,
      "step": 38
    },
    {
      "epoch": 0.013455235466620665,
      "grad_norm": 15.697182655334473,
      "learning_rate": 7.878e-05,
      "loss": 9.885,
      "step": 39
    },
    {
      "epoch": 0.013800241504226323,
      "grad_norm": 15.463223457336426,
      "learning_rate": 8.080000000000001e-05,
      "loss": 8.683,
      "step": 40
    },
    {
      "epoch": 0.014145247541831982,
      "grad_norm": 14.961134910583496,
      "learning_rate": 8.282e-05,
      "loss": 9.5578,
      "step": 41
    },
    {
      "epoch": 0.01449025357943764,
      "grad_norm": 16.346773147583008,
      "learning_rate": 8.484e-05,
      "loss": 8.5754,
      "step": 42
    },
    {
      "epoch": 0.014835259617043298,
      "grad_norm": 16.894882202148438,
      "learning_rate": 8.686e-05,
      "loss": 9.6622,
      "step": 43
    },
    {
      "epoch": 0.015180265654648957,
      "grad_norm": 14.938036918640137,
      "learning_rate": 8.888e-05,
      "loss": 9.8029,
      "step": 44
    },
    {
      "epoch": 0.015525271692254615,
      "grad_norm": 15.628216743469238,
      "learning_rate": 9.09e-05,
      "loss": 9.3792,
      "step": 45
    },
    {
      "epoch": 0.01587027772986027,
      "grad_norm": 16.276512145996094,
      "learning_rate": 9.292000000000001e-05,
      "loss": 8.5939,
      "step": 46
    },
    {
      "epoch": 0.01621528376746593,
      "grad_norm": 15.280430793762207,
      "learning_rate": 9.494e-05,
      "loss": 7.9766,
      "step": 47
    },
    {
      "epoch": 0.016560289805071588,
      "grad_norm": 14.437676429748535,
      "learning_rate": 9.696e-05,
      "loss": 8.8341,
      "step": 48
    },
    {
      "epoch": 0.016905295842677248,
      "grad_norm": 15.67601203918457,
      "learning_rate": 9.898e-05,
      "loss": 9.601,
      "step": 49
    },
    {
      "epoch": 0.017250301880282905,
      "grad_norm": 16.0819034576416,
      "learning_rate": 0.000101,
      "loss": 9.6643,
      "step": 50
    },
    {
      "epoch": 0.017595307917888565,
      "grad_norm": 13.843536376953125,
      "learning_rate": 0.00010302,
      "loss": 8.734,
      "step": 51
    },
    {
      "epoch": 0.01794031395549422,
      "grad_norm": 15.905147552490234,
      "learning_rate": 0.00010504000000000001,
      "loss": 8.5386,
      "step": 52
    },
    {
      "epoch": 0.018285319993099878,
      "grad_norm": 16.874813079833984,
      "learning_rate": 0.00010706000000000001,
      "loss": 9.5245,
      "step": 53
    },
    {
      "epoch": 0.018630326030705538,
      "grad_norm": 14.517797470092773,
      "learning_rate": 0.00010908,
      "loss": 9.8307,
      "step": 54
    },
    {
      "epoch": 0.018975332068311195,
      "grad_norm": 14.526803970336914,
      "learning_rate": 0.00011110000000000002,
      "loss": 8.6252,
      "step": 55
    },
    {
      "epoch": 0.019320338105916855,
      "grad_norm": 14.925496101379395,
      "learning_rate": 0.00011312000000000001,
      "loss": 9.8162,
      "step": 56
    },
    {
      "epoch": 0.01966534414352251,
      "grad_norm": 14.247793197631836,
      "learning_rate": 0.00011514,
      "loss": 9.0644,
      "step": 57
    },
    {
      "epoch": 0.02001035018112817,
      "grad_norm": 15.420412063598633,
      "learning_rate": 0.00011716,
      "loss": 8.8965,
      "step": 58
    },
    {
      "epoch": 0.020355356218733828,
      "grad_norm": 14.702193260192871,
      "learning_rate": 0.00011918,
      "loss": 9.3258,
      "step": 59
    },
    {
      "epoch": 0.020700362256339484,
      "grad_norm": 17.847402572631836,
      "learning_rate": 0.00012119999999999999,
      "loss": 9.2991,
      "step": 60
    },
    {
      "epoch": 0.021045368293945144,
      "grad_norm": 14.710212707519531,
      "learning_rate": 0.00012322,
      "loss": 9.7498,
      "step": 61
    },
    {
      "epoch": 0.0213903743315508,
      "grad_norm": 14.10651683807373,
      "learning_rate": 0.00012524000000000001,
      "loss": 9.0757,
      "step": 62
    },
    {
      "epoch": 0.02173538036915646,
      "grad_norm": 14.680152893066406,
      "learning_rate": 0.00012726,
      "loss": 9.3161,
      "step": 63
    },
    {
      "epoch": 0.022080386406762118,
      "grad_norm": 14.384641647338867,
      "learning_rate": 0.00012928,
      "loss": 9.6402,
      "step": 64
    },
    {
      "epoch": 0.022425392444367778,
      "grad_norm": 13.639796257019043,
      "learning_rate": 0.00013130000000000002,
      "loss": 9.6153,
      "step": 65
    },
    {
      "epoch": 0.022770398481973434,
      "grad_norm": 12.708890914916992,
      "learning_rate": 0.00013332,
      "loss": 8.9354,
      "step": 66
    },
    {
      "epoch": 0.023115404519579094,
      "grad_norm": 16.343677520751953,
      "learning_rate": 0.00013534000000000002,
      "loss": 9.8491,
      "step": 67
    },
    {
      "epoch": 0.02346041055718475,
      "grad_norm": 14.826288223266602,
      "learning_rate": 0.00013736,
      "loss": 9.7392,
      "step": 68
    },
    {
      "epoch": 0.023805416594790407,
      "grad_norm": 14.987595558166504,
      "learning_rate": 0.00013937999999999998,
      "loss": 8.6508,
      "step": 69
    },
    {
      "epoch": 0.024150422632396067,
      "grad_norm": 13.642684936523438,
      "learning_rate": 0.0001414,
      "loss": 8.6308,
      "step": 70
    },
    {
      "epoch": 0.024495428670001724,
      "grad_norm": 15.190869331359863,
      "learning_rate": 0.00014342,
      "loss": 8.5476,
      "step": 71
    },
    {
      "epoch": 0.024840434707607384,
      "grad_norm": 13.480669021606445,
      "learning_rate": 0.00014544,
      "loss": 9.0519,
      "step": 72
    },
    {
      "epoch": 0.02518544074521304,
      "grad_norm": 15.2877779006958,
      "learning_rate": 0.00014746,
      "loss": 9.0994,
      "step": 73
    },
    {
      "epoch": 0.0255304467828187,
      "grad_norm": 13.088419914245605,
      "learning_rate": 0.00014948,
      "loss": 8.6048,
      "step": 74
    },
    {
      "epoch": 0.025875452820424357,
      "grad_norm": 13.846434593200684,
      "learning_rate": 0.0001515,
      "loss": 9.4681,
      "step": 75
    },
    {
      "epoch": 0.026220458858030017,
      "grad_norm": 15.168274879455566,
      "learning_rate": 0.00015352,
      "loss": 9.4533,
      "step": 76
    },
    {
      "epoch": 0.026565464895635674,
      "grad_norm": 12.967020988464355,
      "learning_rate": 0.00015554000000000002,
      "loss": 9.3244,
      "step": 77
    },
    {
      "epoch": 0.02691047093324133,
      "grad_norm": 13.676380157470703,
      "learning_rate": 0.00015756,
      "loss": 8.5062,
      "step": 78
    },
    {
      "epoch": 0.02725547697084699,
      "grad_norm": 12.101982116699219,
      "learning_rate": 0.00015958000000000001,
      "loss": 8.8507,
      "step": 79
    },
    {
      "epoch": 0.027600483008452647,
      "grad_norm": 11.944432258605957,
      "learning_rate": 0.00016160000000000002,
      "loss": 8.8771,
      "step": 80
    },
    {
      "epoch": 0.027945489046058307,
      "grad_norm": 14.021061897277832,
      "learning_rate": 0.00016362,
      "loss": 9.1861,
      "step": 81
    },
    {
      "epoch": 0.028290495083663963,
      "grad_norm": 13.706653594970703,
      "learning_rate": 0.00016564,
      "loss": 9.5414,
      "step": 82
    },
    {
      "epoch": 0.028635501121269624,
      "grad_norm": 15.099771499633789,
      "learning_rate": 0.00016766,
      "loss": 10.0574,
      "step": 83
    },
    {
      "epoch": 0.02898050715887528,
      "grad_norm": 15.680136680603027,
      "learning_rate": 0.00016968,
      "loss": 9.6518,
      "step": 84
    },
    {
      "epoch": 0.02932551319648094,
      "grad_norm": 13.389763832092285,
      "learning_rate": 0.0001717,
      "loss": 9.3081,
      "step": 85
    },
    {
      "epoch": 0.029670519234086597,
      "grad_norm": 11.777738571166992,
      "learning_rate": 0.00017372,
      "loss": 8.3153,
      "step": 86
    },
    {
      "epoch": 0.030015525271692253,
      "grad_norm": 14.047492980957031,
      "learning_rate": 0.00017574,
      "loss": 9.5127,
      "step": 87
    },
    {
      "epoch": 0.030360531309297913,
      "grad_norm": 14.047000885009766,
      "learning_rate": 0.00017776,
      "loss": 9.6181,
      "step": 88
    },
    {
      "epoch": 0.03070553734690357,
      "grad_norm": 13.287646293640137,
      "learning_rate": 0.00017978000000000002,
      "loss": 9.2767,
      "step": 89
    },
    {
      "epoch": 0.03105054338450923,
      "grad_norm": 14.720637321472168,
      "learning_rate": 0.0001818,
      "loss": 9.8054,
      "step": 90
    },
    {
      "epoch": 0.031395549422114886,
      "grad_norm": 16.120899200439453,
      "learning_rate": 0.00018382,
      "loss": 9.4354,
      "step": 91
    },
    {
      "epoch": 0.03174055545972054,
      "grad_norm": 14.359231948852539,
      "learning_rate": 0.00018584000000000002,
      "loss": 8.4884,
      "step": 92
    },
    {
      "epoch": 0.03208556149732621,
      "grad_norm": 15.289236068725586,
      "learning_rate": 0.00018786,
      "loss": 9.751,
      "step": 93
    },
    {
      "epoch": 0.03243056753493186,
      "grad_norm": 14.080328941345215,
      "learning_rate": 0.00018988,
      "loss": 9.3537,
      "step": 94
    },
    {
      "epoch": 0.03277557357253752,
      "grad_norm": 13.745378494262695,
      "learning_rate": 0.0001919,
      "loss": 8.9503,
      "step": 95
    },
    {
      "epoch": 0.033120579610143176,
      "grad_norm": 14.162373542785645,
      "learning_rate": 0.00019392,
      "loss": 9.8135,
      "step": 96
    },
    {
      "epoch": 0.03346558564774883,
      "grad_norm": 13.168984413146973,
      "learning_rate": 0.00019594,
      "loss": 8.7411,
      "step": 97
    },
    {
      "epoch": 0.033810591685354496,
      "grad_norm": 13.539119720458984,
      "learning_rate": 0.00019796,
      "loss": 9.4751,
      "step": 98
    },
    {
      "epoch": 0.03415559772296015,
      "grad_norm": 13.424215316772461,
      "learning_rate": 0.00019998,
      "loss": 8.6123,
      "step": 99
    },
    {
      "epoch": 0.03450060376056581,
      "grad_norm": 12.19062614440918,
      "learning_rate": 0.000202,
      "loss": 8.9604,
      "step": 100
    },
    {
      "epoch": 0.034845609798171466,
      "grad_norm": 14.277666091918945,
      "learning_rate": 0.00020199688492212377,
      "loss": 8.5414,
      "step": 101
    },
    {
      "epoch": 0.03519061583577713,
      "grad_norm": 13.869349479675293,
      "learning_rate": 0.00020198753988064772,
      "loss": 9.6528,
      "step": 102
    },
    {
      "epoch": 0.035535621873382786,
      "grad_norm": 14.230924606323242,
      "learning_rate": 0.00020197196545201806,
      "loss": 8.9774,
      "step": 103
    },
    {
      "epoch": 0.03588062791098844,
      "grad_norm": 15.063007354736328,
      "learning_rate": 0.0002019501625969389,
      "loss": 8.872,
      "step": 104
    },
    {
      "epoch": 0.0362256339485941,
      "grad_norm": 14.859697341918945,
      "learning_rate": 0.00020192213266031304,
      "loss": 8.3321,
      "step": 105
    },
    {
      "epoch": 0.036570639986199756,
      "grad_norm": 13.196431159973145,
      "learning_rate": 0.00020188787737115897,
      "loss": 9.3483,
      "step": 106
    },
    {
      "epoch": 0.03691564602380542,
      "grad_norm": 14.465513229370117,
      "learning_rate": 0.00020184739884250436,
      "loss": 9.5213,
      "step": 107
    },
    {
      "epoch": 0.037260652061411076,
      "grad_norm": 14.234845161437988,
      "learning_rate": 0.00020180069957125544,
      "loss": 8.8897,
      "step": 108
    },
    {
      "epoch": 0.03760565809901673,
      "grad_norm": 15.473494529724121,
      "learning_rate": 0.0002017477824380433,
      "loss": 8.8824,
      "step": 109
    },
    {
      "epoch": 0.03795066413662239,
      "grad_norm": 13.990817070007324,
      "learning_rate": 0.00020168865070704594,
      "loss": 9.1959,
      "step": 110
    },
    {
      "epoch": 0.038295670174228046,
      "grad_norm": 12.264822959899902,
      "learning_rate": 0.00020162330802578706,
      "loss": 8.8838,
      "step": 111
    },
    {
      "epoch": 0.03864067621183371,
      "grad_norm": 12.853663444519043,
      "learning_rate": 0.00020155175842491107,
      "loss": 9.308,
      "step": 112
    },
    {
      "epoch": 0.038985682249439366,
      "grad_norm": 13.567692756652832,
      "learning_rate": 0.0002014740063179344,
      "loss": 9.6124,
      "step": 113
    },
    {
      "epoch": 0.03933068828704502,
      "grad_norm": 12.146742820739746,
      "learning_rate": 0.00020139005650097317,
      "loss": 8.5251,
      "step": 114
    },
    {
      "epoch": 0.03967569432465068,
      "grad_norm": 13.828983306884766,
      "learning_rate": 0.00020129991415244756,
      "loss": 9.8101,
      "step": 115
    },
    {
      "epoch": 0.04002070036225634,
      "grad_norm": 13.260955810546875,
      "learning_rate": 0.00020120358483276227,
      "loss": 9.801,
      "step": 116
    },
    {
      "epoch": 0.040365706399862,
      "grad_norm": 14.162537574768066,
      "learning_rate": 0.00020110107448396346,
      "loss": 10.5028,
      "step": 117
    },
    {
      "epoch": 0.040710712437467655,
      "grad_norm": 12.911920547485352,
      "learning_rate": 0.0002009923894293723,
      "loss": 8.9014,
      "step": 118
    },
    {
      "epoch": 0.04105571847507331,
      "grad_norm": 13.193029403686523,
      "learning_rate": 0.00020087753637319499,
      "loss": 8.9172,
      "step": 119
    },
    {
      "epoch": 0.04140072451267897,
      "grad_norm": 14.418041229248047,
      "learning_rate": 0.00020075652240010892,
      "loss": 10.0696,
      "step": 120
    },
    {
      "epoch": 0.04174573055028463,
      "grad_norm": 14.51584243774414,
      "learning_rate": 0.00020062935497482606,
      "loss": 9.9586,
      "step": 121
    },
    {
      "epoch": 0.04209073658789029,
      "grad_norm": 13.910714149475098,
      "learning_rate": 0.00020049604194163217,
      "loss": 9.1037,
      "step": 122
    },
    {
      "epoch": 0.042435742625495945,
      "grad_norm": 15.043135643005371,
      "learning_rate": 0.00020035659152390313,
      "loss": 9.9946,
      "step": 123
    },
    {
      "epoch": 0.0427807486631016,
      "grad_norm": 14.11701488494873,
      "learning_rate": 0.00020021101232359757,
      "loss": 9.6187,
      "step": 124
    },
    {
      "epoch": 0.043125754700707265,
      "grad_norm": 12.703479766845703,
      "learning_rate": 0.0002000593133207263,
      "loss": 9.2149,
      "step": 125
    },
    {
      "epoch": 0.04347076073831292,
      "grad_norm": 14.16472053527832,
      "learning_rate": 0.00019990150387279835,
      "loss": 8.6816,
      "step": 126
    },
    {
      "epoch": 0.04381576677591858,
      "grad_norm": 17.1020450592041,
      "learning_rate": 0.00019973759371424388,
      "loss": 9.8272,
      "step": 127
    },
    {
      "epoch": 0.044160772813524235,
      "grad_norm": 15.889808654785156,
      "learning_rate": 0.0001995675929558135,
      "loss": 9.5001,
      "step": 128
    },
    {
      "epoch": 0.04450577885112989,
      "grad_norm": 14.733025550842285,
      "learning_rate": 0.0001993915120839548,
      "loss": 8.5339,
      "step": 129
    },
    {
      "epoch": 0.044850784888735555,
      "grad_norm": 16.015302658081055,
      "learning_rate": 0.00019920936196016534,
      "loss": 10.0307,
      "step": 130
    },
    {
      "epoch": 0.04519579092634121,
      "grad_norm": 14.232752799987793,
      "learning_rate": 0.0001990211538203228,
      "loss": 8.9454,
      "step": 131
    },
    {
      "epoch": 0.04554079696394687,
      "grad_norm": 34.98522186279297,
      "learning_rate": 0.00019882689927399174,
      "loss": 9.7892,
      "step": 132
    },
    {
      "epoch": 0.045885803001552525,
      "grad_norm": 14.725144386291504,
      "learning_rate": 0.00019862661030370764,
      "loss": 8.7958,
      "step": 133
    },
    {
      "epoch": 0.04623080903915819,
      "grad_norm": 12.085063934326172,
      "learning_rate": 0.00019842029926423762,
      "loss": 9.7706,
      "step": 134
    },
    {
      "epoch": 0.046575815076763845,
      "grad_norm": 13.047974586486816,
      "learning_rate": 0.00019820797888181837,
      "loss": 9.6162,
      "step": 135
    },
    {
      "epoch": 0.0469208211143695,
      "grad_norm": 11.88274097442627,
      "learning_rate": 0.00019798966225337126,
      "loss": 9.718,
      "step": 136
    },
    {
      "epoch": 0.04726582715197516,
      "grad_norm": 12.723931312561035,
      "learning_rate": 0.00019776536284569425,
      "loss": 10.0413,
      "step": 137
    },
    {
      "epoch": 0.047610833189580815,
      "grad_norm": 12.89576244354248,
      "learning_rate": 0.00019753509449463134,
      "loss": 9.3311,
      "step": 138
    },
    {
      "epoch": 0.04795583922718648,
      "grad_norm": 12.134483337402344,
      "learning_rate": 0.00019729887140421912,
      "loss": 8.4829,
      "step": 139
    },
    {
      "epoch": 0.048300845264792135,
      "grad_norm": 12.476184844970703,
      "learning_rate": 0.00019705670814581052,
      "loss": 8.3054,
      "step": 140
    },
    {
      "epoch": 0.04864585130239779,
      "grad_norm": 16.147558212280273,
      "learning_rate": 0.00019680861965717597,
      "loss": 9.7445,
      "step": 141
    },
    {
      "epoch": 0.04899085734000345,
      "grad_norm": 15.056818008422852,
      "learning_rate": 0.0001965546212415821,
      "loss": 8.921,
      "step": 142
    },
    {
      "epoch": 0.04933586337760911,
      "grad_norm": 14.562697410583496,
      "learning_rate": 0.00019629472856684755,
      "loss": 8.9393,
      "step": 143
    },
    {
      "epoch": 0.04968086941521477,
      "grad_norm": 14.759387016296387,
      "learning_rate": 0.00019602895766437678,
      "loss": 8.4352,
      "step": 144
    },
    {
      "epoch": 0.050025875452820424,
      "grad_norm": 14.771378517150879,
      "learning_rate": 0.00019575732492817092,
      "loss": 9.4725,
      "step": 145
    },
    {
      "epoch": 0.05037088149042608,
      "grad_norm": 13.011505126953125,
      "learning_rate": 0.00019547984711381662,
      "loss": 9.6904,
      "step": 146
    },
    {
      "epoch": 0.05071588752803174,
      "grad_norm": 85.98946380615234,
      "learning_rate": 0.0001951965413374525,
      "loss": 9.2684,
      "step": 147
    },
    {
      "epoch": 0.0510608935656374,
      "grad_norm": 11.127753257751465,
      "learning_rate": 0.00019490742507471338,
      "loss": 9.8473,
      "step": 148
    },
    {
      "epoch": 0.05140589960324306,
      "grad_norm": 11.534029960632324,
      "learning_rate": 0.0001946125161596522,
      "loss": 8.8092,
      "step": 149
    },
    {
      "epoch": 0.051750905640848714,
      "grad_norm": 12.19753646850586,
      "learning_rate": 0.00019431183278363997,
      "loss": 8.5659,
      "step": 150
    },
    {
      "epoch": 0.05209591167845437,
      "grad_norm": 13.595198631286621,
      "learning_rate": 0.00019400539349424367,
      "loss": 9.3434,
      "step": 151
    },
    {
      "epoch": 0.052440917716060034,
      "grad_norm": 14.145285606384277,
      "learning_rate": 0.0001936932171940821,
      "loss": 9.2882,
      "step": 152
    },
    {
      "epoch": 0.05278592375366569,
      "grad_norm": 15.676530838012695,
      "learning_rate": 0.00019337532313966,
      "loss": 9.6214,
      "step": 153
    },
    {
      "epoch": 0.05313092979127135,
      "grad_norm": 13.503463745117188,
      "learning_rate": 0.00019305173094017996,
      "loss": 9.5266,
      "step": 154
    },
    {
      "epoch": 0.053475935828877004,
      "grad_norm": 15.867369651794434,
      "learning_rate": 0.0001927224605563332,
      "loss": 9.5399,
      "step": 155
    },
    {
      "epoch": 0.05382094186648266,
      "grad_norm": 15.228123664855957,
      "learning_rate": 0.00019238753229906797,
      "loss": 9.2673,
      "step": 156
    },
    {
      "epoch": 0.054165947904088324,
      "grad_norm": 12.77865982055664,
      "learning_rate": 0.00019204696682833682,
      "loss": 9.6562,
      "step": 157
    },
    {
      "epoch": 0.05451095394169398,
      "grad_norm": 13.972709655761719,
      "learning_rate": 0.00019170078515182216,
      "loss": 10.269,
      "step": 158
    },
    {
      "epoch": 0.05485595997929964,
      "grad_norm": 13.378512382507324,
      "learning_rate": 0.00019134900862364054,
      "loss": 9.2507,
      "step": 159
    },
    {
      "epoch": 0.055200966016905294,
      "grad_norm": 12.486189842224121,
      "learning_rate": 0.00019099165894302515,
      "loss": 9.6831,
      "step": 160
    },
    {
      "epoch": 0.05554597205451096,
      "grad_norm": 11.630008697509766,
      "learning_rate": 0.00019062875815298763,
      "loss": 9.3057,
      "step": 161
    },
    {
      "epoch": 0.055890978092116614,
      "grad_norm": 13.363422393798828,
      "learning_rate": 0.00019026032863895805,
      "loss": 9.9519,
      "step": 162
    },
    {
      "epoch": 0.05623598412972227,
      "grad_norm": 12.590592384338379,
      "learning_rate": 0.00018988639312740433,
      "loss": 9.1469,
      "step": 163
    },
    {
      "epoch": 0.05658099016732793,
      "grad_norm": 13.203914642333984,
      "learning_rate": 0.0001895069746844302,
      "loss": 9.3678,
      "step": 164
    },
    {
      "epoch": 0.056925996204933584,
      "grad_norm": 14.903193473815918,
      "learning_rate": 0.00018912209671435252,
      "loss": 9.119,
      "step": 165
    },
    {
      "epoch": 0.05727100224253925,
      "grad_norm": 14.591456413269043,
      "learning_rate": 0.00018873178295825732,
      "loss": 9.3862,
      "step": 166
    },
    {
      "epoch": 0.057616008280144904,
      "grad_norm": 13.231369018554688,
      "learning_rate": 0.00018833605749253566,
      "loss": 9.5783,
      "step": 167
    },
    {
      "epoch": 0.05796101431775056,
      "grad_norm": 14.425864219665527,
      "learning_rate": 0.00018793494472739831,
      "loss": 9.827,
      "step": 168
    },
    {
      "epoch": 0.05830602035535622,
      "grad_norm": 16.900150299072266,
      "learning_rate": 0.00018752846940537003,
      "loss": 10.5952,
      "step": 169
    },
    {
      "epoch": 0.05865102639296188,
      "grad_norm": 14.382865905761719,
      "learning_rate": 0.0001871166565997633,
      "loss": 10.2463,
      "step": 170
    },
    {
      "epoch": 0.05899603243056754,
      "grad_norm": 14.044133186340332,
      "learning_rate": 0.00018669953171313188,
      "loss": 9.1259,
      "step": 171
    },
    {
      "epoch": 0.05934103846817319,
      "grad_norm": 15.003429412841797,
      "learning_rate": 0.00018627712047570352,
      "loss": 10.78,
      "step": 172
    },
    {
      "epoch": 0.05968604450577885,
      "grad_norm": 15.312614440917969,
      "learning_rate": 0.0001858494489437931,
      "loss": 9.4416,
      "step": 173
    },
    {
      "epoch": 0.060031050543384507,
      "grad_norm": 12.405970573425293,
      "learning_rate": 0.0001854165434981953,
      "loss": 9.1612,
      "step": 174
    },
    {
      "epoch": 0.06037605658099017,
      "grad_norm": 13.486007690429688,
      "learning_rate": 0.00018497843084255708,
      "loss": 9.6598,
      "step": 175
    },
    {
      "epoch": 0.06072106261859583,
      "grad_norm": 14.553987503051758,
      "learning_rate": 0.00018453513800173072,
      "loss": 8.7911,
      "step": 176
    },
    {
      "epoch": 0.06106606865620148,
      "grad_norm": 14.363768577575684,
      "learning_rate": 0.00018408669232010684,
      "loss": 9.8944,
      "step": 177
    },
    {
      "epoch": 0.06141107469380714,
      "grad_norm": 13.335902214050293,
      "learning_rate": 0.00018363312145992737,
      "loss": 10.2723,
      "step": 178
    },
    {
      "epoch": 0.061756080731412796,
      "grad_norm": 12.759366989135742,
      "learning_rate": 0.0001831744533995795,
      "loss": 8.7982,
      "step": 179
    },
    {
      "epoch": 0.06210108676901846,
      "grad_norm": 13.948934555053711,
      "learning_rate": 0.00018271071643186968,
      "loss": 8.9383,
      "step": 180
    },
    {
      "epoch": 0.062446092806624116,
      "grad_norm": 14.048227310180664,
      "learning_rate": 0.00018224193916227852,
      "loss": 9.2508,
      "step": 181
    },
    {
      "epoch": 0.06279109884422977,
      "grad_norm": 14.153533935546875,
      "learning_rate": 0.00018176815050719615,
      "loss": 8.811,
      "step": 182
    },
    {
      "epoch": 0.06313610488183544,
      "grad_norm": 13.415421485900879,
      "learning_rate": 0.00018128937969213852,
      "loss": 9.4737,
      "step": 183
    },
    {
      "epoch": 0.06348111091944109,
      "grad_norm": 12.875298500061035,
      "learning_rate": 0.00018080565624994474,
      "loss": 9.6055,
      "step": 184
    },
    {
      "epoch": 0.06382611695704675,
      "grad_norm": 14.840189933776855,
      "learning_rate": 0.00018031701001895524,
      "loss": 9.5473,
      "step": 185
    },
    {
      "epoch": 0.06417112299465241,
      "grad_norm": 14.026057243347168,
      "learning_rate": 0.0001798234711411713,
      "loss": 9.3531,
      "step": 186
    },
    {
      "epoch": 0.06451612903225806,
      "grad_norm": 17.18976402282715,
      "learning_rate": 0.00017932507006039567,
      "loss": 9.5941,
      "step": 187
    },
    {
      "epoch": 0.06486113506986373,
      "grad_norm": 14.190940856933594,
      "learning_rate": 0.0001788218375203547,
      "loss": 10.1677,
      "step": 188
    },
    {
      "epoch": 0.06520614110746938,
      "grad_norm": 12.98277473449707,
      "learning_rate": 0.00017831380456280192,
      "loss": 8.8519,
      "step": 189
    },
    {
      "epoch": 0.06555114714507504,
      "grad_norm": 12.706244468688965,
      "learning_rate": 0.00017780100252560313,
      "loss": 9.4913,
      "step": 190
    },
    {
      "epoch": 0.0658961531826807,
      "grad_norm": 14.777141571044922,
      "learning_rate": 0.00017728346304080357,
      "loss": 9.5007,
      "step": 191
    },
    {
      "epoch": 0.06624115922028635,
      "grad_norm": 12.656100273132324,
      "learning_rate": 0.0001767612180326764,
      "loss": 9.597,
      "step": 192
    },
    {
      "epoch": 0.06658616525789202,
      "grad_norm": 12.695791244506836,
      "learning_rate": 0.00017623429971575384,
      "loss": 8.5745,
      "step": 193
    },
    {
      "epoch": 0.06693117129549767,
      "grad_norm": 13.431217193603516,
      "learning_rate": 0.0001757027405928396,
      "loss": 9.4003,
      "step": 194
    },
    {
      "epoch": 0.06727617733310333,
      "grad_norm": 11.731524467468262,
      "learning_rate": 0.00017516657345300425,
      "loss": 8.9719,
      "step": 195
    },
    {
      "epoch": 0.06762118337070899,
      "grad_norm": 17.55036163330078,
      "learning_rate": 0.00017462583136956258,
      "loss": 9.89,
      "step": 196
    },
    {
      "epoch": 0.06796618940831464,
      "grad_norm": 11.960362434387207,
      "learning_rate": 0.00017408054769803337,
      "loss": 9.223,
      "step": 197
    },
    {
      "epoch": 0.0683111954459203,
      "grad_norm": 13.012106895446777,
      "learning_rate": 0.00017353075607408209,
      "loss": 9.2416,
      "step": 198
    },
    {
      "epoch": 0.06865620148352596,
      "grad_norm": 13.101168632507324,
      "learning_rate": 0.00017297649041144575,
      "loss": 10.0126,
      "step": 199
    },
    {
      "epoch": 0.06900120752113162,
      "grad_norm": 12.447086334228516,
      "learning_rate": 0.0001724177848998413,
      "loss": 10.7401,
      "step": 200
    },
    {
      "epoch": 0.06934621355873728,
      "grad_norm": 13.678348541259766,
      "learning_rate": 0.00017185467400285644,
      "loss": 9.8607,
      "step": 201
    },
    {
      "epoch": 0.06969121959634293,
      "grad_norm": 11.554344177246094,
      "learning_rate": 0.00017128719245582374,
      "loss": 9.3625,
      "step": 202
    },
    {
      "epoch": 0.0700362256339486,
      "grad_norm": 13.678261756896973,
      "learning_rate": 0.00017071537526367817,
      "loss": 10.6498,
      "step": 203
    },
    {
      "epoch": 0.07038123167155426,
      "grad_norm": 12.270160675048828,
      "learning_rate": 0.00017013925769879755,
      "loss": 8.3938,
      "step": 204
    },
    {
      "epoch": 0.07072623770915991,
      "grad_norm": 12.534881591796875,
      "learning_rate": 0.00016955887529882714,
      "loss": 9.2155,
      "step": 205
    },
    {
      "epoch": 0.07107124374676557,
      "grad_norm": 12.138201713562012,
      "learning_rate": 0.0001689742638644871,
      "loss": 9.4221,
      "step": 206
    },
    {
      "epoch": 0.07141624978437122,
      "grad_norm": 11.361064910888672,
      "learning_rate": 0.00016838545945736458,
      "loss": 8.919,
      "step": 207
    },
    {
      "epoch": 0.07176125582197689,
      "grad_norm": 12.660870552062988,
      "learning_rate": 0.00016779249839768884,
      "loss": 8.3665,
      "step": 208
    },
    {
      "epoch": 0.07210626185958255,
      "grad_norm": 14.647711753845215,
      "learning_rate": 0.00016719541726209117,
      "loss": 10.0559,
      "step": 209
    },
    {
      "epoch": 0.0724512678971882,
      "grad_norm": 13.76112174987793,
      "learning_rate": 0.00016659425288134854,
      "loss": 9.7268,
      "step": 210
    },
    {
      "epoch": 0.07279627393479386,
      "grad_norm": 13.205056190490723,
      "learning_rate": 0.00016598904233811168,
      "loss": 9.7256,
      "step": 211
    },
    {
      "epoch": 0.07314127997239951,
      "grad_norm": 12.077413558959961,
      "learning_rate": 0.00016537982296461768,
      "loss": 8.3806,
      "step": 212
    },
    {
      "epoch": 0.07348628601000518,
      "grad_norm": 13.776256561279297,
      "learning_rate": 0.00016476663234038717,
      "loss": 9.8824,
      "step": 213
    },
    {
      "epoch": 0.07383129204761084,
      "grad_norm": 13.2633638381958,
      "learning_rate": 0.00016414950828990625,
      "loss": 8.9381,
      "step": 214
    },
    {
      "epoch": 0.07417629808521649,
      "grad_norm": 12.64013957977295,
      "learning_rate": 0.00016352848888029326,
      "loss": 8.5029,
      "step": 215
    },
    {
      "epoch": 0.07452130412282215,
      "grad_norm": 13.421082496643066,
      "learning_rate": 0.00016290361241895064,
      "loss": 8.3436,
      "step": 216
    },
    {
      "epoch": 0.0748663101604278,
      "grad_norm": 12.582505226135254,
      "learning_rate": 0.00016227491745120196,
      "loss": 8.5146,
      "step": 217
    },
    {
      "epoch": 0.07521131619803346,
      "grad_norm": 15.521428108215332,
      "learning_rate": 0.0001616424427579143,
      "loss": 9.4144,
      "step": 218
    },
    {
      "epoch": 0.07555632223563913,
      "grad_norm": 16.33211326599121,
      "learning_rate": 0.0001610062273531059,
      "loss": 9.9581,
      "step": 219
    },
    {
      "epoch": 0.07590132827324478,
      "grad_norm": 13.193209648132324,
      "learning_rate": 0.00016036631048153979,
      "loss": 8.6518,
      "step": 220
    },
    {
      "epoch": 0.07624633431085044,
      "grad_norm": 14.05943775177002,
      "learning_rate": 0.0001597227316163029,
      "loss": 9.0778,
      "step": 221
    },
    {
      "epoch": 0.07659134034845609,
      "grad_norm": 12.811190605163574,
      "learning_rate": 0.00015907553045637116,
      "loss": 9.5572,
      "step": 222
    },
    {
      "epoch": 0.07693634638606175,
      "grad_norm": 12.563728332519531,
      "learning_rate": 0.00015842474692416068,
      "loss": 8.817,
      "step": 223
    },
    {
      "epoch": 0.07728135242366742,
      "grad_norm": 11.644241333007812,
      "learning_rate": 0.0001577704211630652,
      "loss": 8.483,
      "step": 224
    },
    {
      "epoch": 0.07762635846127307,
      "grad_norm": 12.451133728027344,
      "learning_rate": 0.00015711259353497981,
      "loss": 9.6708,
      "step": 225
    },
    {
      "epoch": 0.07797136449887873,
      "grad_norm": 11.948263168334961,
      "learning_rate": 0.0001564513046178113,
      "loss": 9.2661,
      "step": 226
    },
    {
      "epoch": 0.0783163705364844,
      "grad_norm": 12.96564769744873,
      "learning_rate": 0.000155786595202975,
      "loss": 10.571,
      "step": 227
    },
    {
      "epoch": 0.07866137657409004,
      "grad_norm": 13.042709350585938,
      "learning_rate": 0.00015511850629287865,
      "loss": 8.1806,
      "step": 228
    },
    {
      "epoch": 0.07900638261169571,
      "grad_norm": 12.97872257232666,
      "learning_rate": 0.00015444707909839325,
      "loss": 8.8924,
      "step": 229
    },
    {
      "epoch": 0.07935138864930136,
      "grad_norm": 14.92879581451416,
      "learning_rate": 0.00015377235503631083,
      "loss": 9.968,
      "step": 230
    },
    {
      "epoch": 0.07969639468690702,
      "grad_norm": 13.545868873596191,
      "learning_rate": 0.0001530943757267898,
      "loss": 9.4662,
      "step": 231
    },
    {
      "epoch": 0.08004140072451268,
      "grad_norm": 12.4180269241333,
      "learning_rate": 0.00015241318299078751,
      "loss": 8.7383,
      "step": 232
    },
    {
      "epoch": 0.08038640676211833,
      "grad_norm": 14.102592468261719,
      "learning_rate": 0.00015172881884748063,
      "loss": 9.5623,
      "step": 233
    },
    {
      "epoch": 0.080731412799724,
      "grad_norm": 13.51630687713623,
      "learning_rate": 0.00015104132551167318,
      "loss": 9.9228,
      "step": 234
    },
    {
      "epoch": 0.08107641883732965,
      "grad_norm": 12.33234691619873,
      "learning_rate": 0.00015035074539119248,
      "loss": 9.5282,
      "step": 235
    },
    {
      "epoch": 0.08142142487493531,
      "grad_norm": 16.598876953125,
      "learning_rate": 0.00014965712108427323,
      "loss": 8.5303,
      "step": 236
    },
    {
      "epoch": 0.08176643091254097,
      "grad_norm": 12.998722076416016,
      "learning_rate": 0.00014896049537693005,
      "loss": 9.2569,
      "step": 237
    },
    {
      "epoch": 0.08211143695014662,
      "grad_norm": 12.277347564697266,
      "learning_rate": 0.00014826091124031792,
      "loss": 9.4612,
      "step": 238
    },
    {
      "epoch": 0.08245644298775229,
      "grad_norm": 12.43691349029541,
      "learning_rate": 0.0001475584118280817,
      "loss": 8.9428,
      "step": 239
    },
    {
      "epoch": 0.08280144902535794,
      "grad_norm": 14.75731086730957,
      "learning_rate": 0.00014685304047369423,
      "loss": 9.0155,
      "step": 240
    },
    {
      "epoch": 0.0831464550629636,
      "grad_norm": 12.837821960449219,
      "learning_rate": 0.00014614484068778324,
      "loss": 8.717,
      "step": 241
    },
    {
      "epoch": 0.08349146110056926,
      "grad_norm": 12.87870979309082,
      "learning_rate": 0.00014543385615544744,
      "loss": 8.7477,
      "step": 242
    },
    {
      "epoch": 0.08383646713817491,
      "grad_norm": 12.544157028198242,
      "learning_rate": 0.00014472013073356184,
      "loss": 10.2256,
      "step": 243
    },
    {
      "epoch": 0.08418147317578058,
      "grad_norm": 12.770133972167969,
      "learning_rate": 0.00014400370844807234,
      "loss": 9.1473,
      "step": 244
    },
    {
      "epoch": 0.08452647921338624,
      "grad_norm": 13.194252014160156,
      "learning_rate": 0.00014328463349128025,
      "loss": 9.2346,
      "step": 245
    },
    {
      "epoch": 0.08487148525099189,
      "grad_norm": 14.135968208312988,
      "learning_rate": 0.000142562950219116,
      "loss": 8.905,
      "step": 246
    },
    {
      "epoch": 0.08521649128859755,
      "grad_norm": 13.687217712402344,
      "learning_rate": 0.00014183870314840325,
      "loss": 9.5711,
      "step": 247
    },
    {
      "epoch": 0.0855614973262032,
      "grad_norm": 13.425223350524902,
      "learning_rate": 0.00014111193695411285,
      "loss": 9.4454,
      "step": 248
    },
    {
      "epoch": 0.08590650336380887,
      "grad_norm": 12.061142921447754,
      "learning_rate": 0.00014038269646660703,
      "loss": 8.922,
      "step": 249
    },
    {
      "epoch": 0.08625150940141453,
      "grad_norm": 12.300885200500488,
      "learning_rate": 0.00013965102666887408,
      "loss": 9.3866,
      "step": 250
    },
    {
      "epoch": 0.08659651543902018,
      "grad_norm": 11.672247886657715,
      "learning_rate": 0.0001389169726937536,
      "loss": 9.1472,
      "step": 251
    },
    {
      "epoch": 0.08694152147662584,
      "grad_norm": 11.093493461608887,
      "learning_rate": 0.0001381805798211525,
      "loss": 9.8978,
      "step": 252
    },
    {
      "epoch": 0.0872865275142315,
      "grad_norm": 11.971585273742676,
      "learning_rate": 0.00013744189347525182,
      "loss": 9.4308,
      "step": 253
    },
    {
      "epoch": 0.08763153355183716,
      "grad_norm": 11.615097045898438,
      "learning_rate": 0.00013670095922170498,
      "loss": 8.5866,
      "step": 254
    },
    {
      "epoch": 0.08797653958944282,
      "grad_norm": 12.57749080657959,
      "learning_rate": 0.00013595782276482678,
      "loss": 8.6278,
      "step": 255
    },
    {
      "epoch": 0.08832154562704847,
      "grad_norm": 12.750253677368164,
      "learning_rate": 0.00013521252994477446,
      "loss": 8.49,
      "step": 256
    },
    {
      "epoch": 0.08866655166465413,
      "grad_norm": 12.982474327087402,
      "learning_rate": 0.00013446512673471965,
      "loss": 9.349,
      "step": 257
    },
    {
      "epoch": 0.08901155770225978,
      "grad_norm": 13.189504623413086,
      "learning_rate": 0.0001337156592380131,
      "loss": 9.6014,
      "step": 258
    },
    {
      "epoch": 0.08935656373986545,
      "grad_norm": 13.597434997558594,
      "learning_rate": 0.0001329641736853402,
      "loss": 8.9983,
      "step": 259
    },
    {
      "epoch": 0.08970156977747111,
      "grad_norm": 12.793052673339844,
      "learning_rate": 0.0001322107164318697,
      "loss": 9.5824,
      "step": 260
    },
    {
      "epoch": 0.09004657581507676,
      "grad_norm": 13.007431983947754,
      "learning_rate": 0.00013145533395439405,
      "loss": 9.0784,
      "step": 261
    },
    {
      "epoch": 0.09039158185268242,
      "grad_norm": 12.095006942749023,
      "learning_rate": 0.0001306980728484627,
      "loss": 8.5778,
      "step": 262
    },
    {
      "epoch": 0.09073658789028809,
      "grad_norm": 13.33928108215332,
      "learning_rate": 0.00012993897982550764,
      "loss": 8.9633,
      "step": 263
    },
    {
      "epoch": 0.09108159392789374,
      "grad_norm": 12.151418685913086,
      "learning_rate": 0.00012917810170996218,
      "loss": 8.1099,
      "step": 264
    },
    {
      "epoch": 0.0914265999654994,
      "grad_norm": 11.10233211517334,
      "learning_rate": 0.0001284154854363725,
      "loss": 8.4968,
      "step": 265
    },
    {
      "epoch": 0.09177160600310505,
      "grad_norm": 12.609495162963867,
      "learning_rate": 0.00012765117804650267,
      "loss": 9.33,
      "step": 266
    },
    {
      "epoch": 0.09211661204071071,
      "grad_norm": 14.439665794372559,
      "learning_rate": 0.00012688522668643268,
      "loss": 9.8422,
      "step": 267
    },
    {
      "epoch": 0.09246161807831638,
      "grad_norm": 13.673382759094238,
      "learning_rate": 0.00012611767860365038,
      "loss": 9.5838,
      "step": 268
    },
    {
      "epoch": 0.09280662411592203,
      "grad_norm": 12.445148468017578,
      "learning_rate": 0.00012534858114413692,
      "loss": 9.117,
      "step": 269
    },
    {
      "epoch": 0.09315163015352769,
      "grad_norm": 14.146788597106934,
      "learning_rate": 0.00012457798174944645,
      "loss": 9.3119,
      "step": 270
    },
    {
      "epoch": 0.09349663619113334,
      "grad_norm": 12.235092163085938,
      "learning_rate": 0.0001238059279537795,
      "loss": 8.747,
      "step": 271
    },
    {
      "epoch": 0.093841642228739,
      "grad_norm": 13.520039558410645,
      "learning_rate": 0.00012303246738105082,
      "loss": 10.3798,
      "step": 272
    },
    {
      "epoch": 0.09418664826634467,
      "grad_norm": 13.536327362060547,
      "learning_rate": 0.00012225764774195186,
      "loss": 8.835,
      "step": 273
    },
    {
      "epoch": 0.09453165430395032,
      "grad_norm": 12.270842552185059,
      "learning_rate": 0.00012148151683100776,
      "loss": 8.9087,
      "step": 274
    },
    {
      "epoch": 0.09487666034155598,
      "grad_norm": 13.040043830871582,
      "learning_rate": 0.00012070412252362897,
      "loss": 8.9546,
      "step": 275
    },
    {
      "epoch": 0.09522166637916163,
      "grad_norm": 12.913331985473633,
      "learning_rate": 0.0001199255127731582,
      "loss": 8.7691,
      "step": 276
    },
    {
      "epoch": 0.09556667241676729,
      "grad_norm": 13.505439758300781,
      "learning_rate": 0.00011914573560791246,
      "loss": 9.9893,
      "step": 277
    },
    {
      "epoch": 0.09591167845437296,
      "grad_norm": 11.862720489501953,
      "learning_rate": 0.00011836483912822035,
      "loss": 8.98,
      "step": 278
    },
    {
      "epoch": 0.0962566844919786,
      "grad_norm": 12.947311401367188,
      "learning_rate": 0.00011758287150345516,
      "loss": 9.2353,
      "step": 279
    },
    {
      "epoch": 0.09660169052958427,
      "grad_norm": 13.35925579071045,
      "learning_rate": 0.00011679988096906333,
      "loss": 9.5359,
      "step": 280
    },
    {
      "epoch": 0.09694669656718992,
      "grad_norm": 12.566567420959473,
      "learning_rate": 0.00011601591582358924,
      "loss": 9.216,
      "step": 281
    },
    {
      "epoch": 0.09729170260479558,
      "grad_norm": 13.095525741577148,
      "learning_rate": 0.00011523102442569585,
      "loss": 10.3427,
      "step": 282
    },
    {
      "epoch": 0.09763670864240125,
      "grad_norm": 11.8966703414917,
      "learning_rate": 0.00011444525519118179,
      "loss": 9.1174,
      "step": 283
    },
    {
      "epoch": 0.0979817146800069,
      "grad_norm": 12.056434631347656,
      "learning_rate": 0.00011365865658999474,
      "loss": 8.4258,
      "step": 284
    },
    {
      "epoch": 0.09832672071761256,
      "grad_norm": 11.04202651977539,
      "learning_rate": 0.00011287127714324162,
      "loss": 9.5051,
      "step": 285
    },
    {
      "epoch": 0.09867172675521822,
      "grad_norm": 12.701394081115723,
      "learning_rate": 0.00011208316542019556,
      "loss": 10.4461,
      "step": 286
    },
    {
      "epoch": 0.09901673279282387,
      "grad_norm": 11.603574752807617,
      "learning_rate": 0.00011129437003530006,
      "loss": 9.326,
      "step": 287
    },
    {
      "epoch": 0.09936173883042954,
      "grad_norm": 11.57833194732666,
      "learning_rate": 0.00011050493964516997,
      "loss": 8.8177,
      "step": 288
    },
    {
      "epoch": 0.09970674486803519,
      "grad_norm": 11.816858291625977,
      "learning_rate": 0.00010971492294559029,
      "loss": 9.1732,
      "step": 289
    },
    {
      "epoch": 0.10005175090564085,
      "grad_norm": 12.905196189880371,
      "learning_rate": 0.00010892436866851235,
      "loss": 9.1141,
      "step": 290
    },
    {
      "epoch": 0.10039675694324651,
      "grad_norm": 14.986270904541016,
      "learning_rate": 0.00010813332557904784,
      "loss": 8.8376,
      "step": 291
    },
    {
      "epoch": 0.10074176298085216,
      "grad_norm": 11.423657417297363,
      "learning_rate": 0.00010734184247246066,
      "loss": 7.5046,
      "step": 292
    },
    {
      "epoch": 0.10108676901845783,
      "grad_norm": 12.934806823730469,
      "learning_rate": 0.00010654996817115704,
      "loss": 8.7451,
      "step": 293
    },
    {
      "epoch": 0.10143177505606348,
      "grad_norm": 13.150904655456543,
      "learning_rate": 0.00010575775152167391,
      "loss": 9.0248,
      "step": 294
    },
    {
      "epoch": 0.10177678109366914,
      "grad_norm": 13.334962844848633,
      "learning_rate": 0.00010496524139166594,
      "loss": 9.3297,
      "step": 295
    },
    {
      "epoch": 0.1021217871312748,
      "grad_norm": 12.290640830993652,
      "learning_rate": 0.00010417248666689095,
      "loss": 8.3023,
      "step": 296
    },
    {
      "epoch": 0.10246679316888045,
      "grad_norm": 11.23175048828125,
      "learning_rate": 0.00010337953624819464,
      "loss": 8.463,
      "step": 297
    },
    {
      "epoch": 0.10281179920648612,
      "grad_norm": 13.800769805908203,
      "learning_rate": 0.0001025864390484939,
      "loss": 9.7216,
      "step": 298
    },
    {
      "epoch": 0.10315680524409176,
      "grad_norm": 12.985389709472656,
      "learning_rate": 0.00010179324398975984,
      "loss": 9.7683,
      "step": 299
    },
    {
      "epoch": 0.10350181128169743,
      "grad_norm": 12.610849380493164,
      "learning_rate": 0.000101,
      "loss": 8.3891,
      "step": 300
    },
    {
      "epoch": 0.10384681731930309,
      "grad_norm": 11.938310623168945,
      "learning_rate": 0.00010020675601024019,
      "loss": 8.587,
      "step": 301
    },
    {
      "epoch": 0.10419182335690874,
      "grad_norm": 11.89841079711914,
      "learning_rate": 9.941356095150613e-05,
      "loss": 8.7931,
      "step": 302
    },
    {
      "epoch": 0.1045368293945144,
      "grad_norm": 12.698895454406738,
      "learning_rate": 9.862046375180539e-05,
      "loss": 8.786,
      "step": 303
    },
    {
      "epoch": 0.10488183543212007,
      "grad_norm": 11.428459167480469,
      "learning_rate": 9.782751333310905e-05,
      "loss": 8.5711,
      "step": 304
    },
    {
      "epoch": 0.10522684146972572,
      "grad_norm": 11.027230262756348,
      "learning_rate": 9.70347586083341e-05,
      "loss": 9.051,
      "step": 305
    },
    {
      "epoch": 0.10557184750733138,
      "grad_norm": 11.26461410522461,
      "learning_rate": 9.62422484783261e-05,
      "loss": 8.4764,
      "step": 306
    },
    {
      "epoch": 0.10591685354493703,
      "grad_norm": 11.53715991973877,
      "learning_rate": 9.5450031828843e-05,
      "loss": 9.6799,
      "step": 307
    },
    {
      "epoch": 0.1062618595825427,
      "grad_norm": 12.850890159606934,
      "learning_rate": 9.465815752753935e-05,
      "loss": 10.62,
      "step": 308
    },
    {
      "epoch": 0.10660686562014836,
      "grad_norm": 11.314481735229492,
      "learning_rate": 9.386667442095219e-05,
      "loss": 9.9539,
      "step": 309
    },
    {
      "epoch": 0.10695187165775401,
      "grad_norm": 10.407557487487793,
      "learning_rate": 9.307563133148767e-05,
      "loss": 8.7503,
      "step": 310
    },
    {
      "epoch": 0.10729687769535967,
      "grad_norm": 12.086105346679688,
      "learning_rate": 9.228507705440976e-05,
      "loss": 8.2982,
      "step": 311
    },
    {
      "epoch": 0.10764188373296532,
      "grad_norm": 11.775453567504883,
      "learning_rate": 9.149506035483005e-05,
      "loss": 9.2283,
      "step": 312
    },
    {
      "epoch": 0.10798688977057098,
      "grad_norm": 11.29387092590332,
      "learning_rate": 9.070562996469997e-05,
      "loss": 9.627,
      "step": 313
    },
    {
      "epoch": 0.10833189580817665,
      "grad_norm": 11.330912590026855,
      "learning_rate": 8.991683457980443e-05,
      "loss": 8.9923,
      "step": 314
    },
    {
      "epoch": 0.1086769018457823,
      "grad_norm": 11.065328598022461,
      "learning_rate": 8.912872285675841e-05,
      "loss": 8.5333,
      "step": 315
    },
    {
      "epoch": 0.10902190788338796,
      "grad_norm": 12.606433868408203,
      "learning_rate": 8.834134341000527e-05,
      "loss": 8.502,
      "step": 316
    },
    {
      "epoch": 0.10936691392099361,
      "grad_norm": 12.832488059997559,
      "learning_rate": 8.755474480881823e-05,
      "loss": 9.3703,
      "step": 317
    },
    {
      "epoch": 0.10971191995859927,
      "grad_norm": 12.532611846923828,
      "learning_rate": 8.676897557430415e-05,
      "loss": 8.8157,
      "step": 318
    },
    {
      "epoch": 0.11005692599620494,
      "grad_norm": 11.844246864318848,
      "learning_rate": 8.598408417641078e-05,
      "loss": 10.0304,
      "step": 319
    },
    {
      "epoch": 0.11040193203381059,
      "grad_norm": 12.978519439697266,
      "learning_rate": 8.520011903093666e-05,
      "loss": 8.9027,
      "step": 320
    },
    {
      "epoch": 0.11074693807141625,
      "grad_norm": 13.221746444702148,
      "learning_rate": 8.441712849654485e-05,
      "loss": 9.0624,
      "step": 321
    },
    {
      "epoch": 0.11109194410902191,
      "grad_norm": 11.086380958557129,
      "learning_rate": 8.363516087177962e-05,
      "loss": 8.672,
      "step": 322
    },
    {
      "epoch": 0.11143695014662756,
      "grad_norm": 12.500789642333984,
      "learning_rate": 8.285426439208755e-05,
      "loss": 9.1655,
      "step": 323
    },
    {
      "epoch": 0.11178195618423323,
      "grad_norm": 12.14895248413086,
      "learning_rate": 8.20744872268418e-05,
      "loss": 9.2437,
      "step": 324
    },
    {
      "epoch": 0.11212696222183888,
      "grad_norm": 11.7655611038208,
      "learning_rate": 8.129587747637105e-05,
      "loss": 9.0607,
      "step": 325
    },
    {
      "epoch": 0.11247196825944454,
      "grad_norm": 11.949422836303711,
      "learning_rate": 8.051848316899227e-05,
      "loss": 9.4289,
      "step": 326
    },
    {
      "epoch": 0.1128169742970502,
      "grad_norm": 11.998117446899414,
      "learning_rate": 7.974235225804814e-05,
      "loss": 9.1564,
      "step": 327
    },
    {
      "epoch": 0.11316198033465585,
      "grad_norm": 12.330270767211914,
      "learning_rate": 7.896753261894923e-05,
      "loss": 10.0202,
      "step": 328
    },
    {
      "epoch": 0.11350698637226152,
      "grad_norm": 11.510910034179688,
      "learning_rate": 7.819407204622054e-05,
      "loss": 8.3607,
      "step": 329
    },
    {
      "epoch": 0.11385199240986717,
      "grad_norm": 12.662656784057617,
      "learning_rate": 7.74220182505536e-05,
      "loss": 9.6127,
      "step": 330
    },
    {
      "epoch": 0.11419699844747283,
      "grad_norm": 11.429125785827637,
      "learning_rate": 7.665141885586312e-05,
      "loss": 8.8394,
      "step": 331
    },
    {
      "epoch": 0.1145420044850785,
      "grad_norm": 10.766180992126465,
      "learning_rate": 7.588232139634968e-05,
      "loss": 8.7825,
      "step": 332
    },
    {
      "epoch": 0.11488701052268414,
      "grad_norm": 12.781359672546387,
      "learning_rate": 7.511477331356733e-05,
      "loss": 8.3732,
      "step": 333
    },
    {
      "epoch": 0.11523201656028981,
      "grad_norm": 11.398038864135742,
      "learning_rate": 7.434882195349736e-05,
      "loss": 8.2819,
      "step": 334
    },
    {
      "epoch": 0.11557702259789546,
      "grad_norm": 11.282587051391602,
      "learning_rate": 7.358451456362751e-05,
      "loss": 8.7989,
      "step": 335
    },
    {
      "epoch": 0.11592202863550112,
      "grad_norm": 11.712827682495117,
      "learning_rate": 7.282189829003785e-05,
      "loss": 8.7423,
      "step": 336
    },
    {
      "epoch": 0.11626703467310678,
      "grad_norm": 11.534231185913086,
      "learning_rate": 7.206102017449237e-05,
      "loss": 8.4394,
      "step": 337
    },
    {
      "epoch": 0.11661204071071243,
      "grad_norm": 10.476648330688477,
      "learning_rate": 7.130192715153731e-05,
      "loss": 8.8547,
      "step": 338
    },
    {
      "epoch": 0.1169570467483181,
      "grad_norm": 13.111579895019531,
      "learning_rate": 7.054466604560595e-05,
      "loss": 10.0197,
      "step": 339
    },
    {
      "epoch": 0.11730205278592376,
      "grad_norm": 12.214439392089844,
      "learning_rate": 6.978928356813031e-05,
      "loss": 8.9993,
      "step": 340
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 12.705840110778809,
      "learning_rate": 6.90358263146598e-05,
      "loss": 7.703,
      "step": 341
    },
    {
      "epoch": 0.11799206486113507,
      "grad_norm": 12.527071952819824,
      "learning_rate": 6.828434076198693e-05,
      "loss": 9.62,
      "step": 342
    },
    {
      "epoch": 0.11833707089874072,
      "grad_norm": 12.797919273376465,
      "learning_rate": 6.753487326528033e-05,
      "loss": 9.4258,
      "step": 343
    },
    {
      "epoch": 0.11868207693634639,
      "grad_norm": 13.112407684326172,
      "learning_rate": 6.678747005522557e-05,
      "loss": 9.3108,
      "step": 344
    },
    {
      "epoch": 0.11902708297395205,
      "grad_norm": 11.739080429077148,
      "learning_rate": 6.60421772351732e-05,
      "loss": 9.0688,
      "step": 345
    },
    {
      "epoch": 0.1193720890115577,
      "grad_norm": 12.013139724731445,
      "learning_rate": 6.529904077829505e-05,
      "loss": 9.2455,
      "step": 346
    },
    {
      "epoch": 0.11971709504916336,
      "grad_norm": 12.952916145324707,
      "learning_rate": 6.455810652474817e-05,
      "loss": 9.7039,
      "step": 347
    },
    {
      "epoch": 0.12006210108676901,
      "grad_norm": 12.223871231079102,
      "learning_rate": 6.381942017884753e-05,
      "loss": 9.1365,
      "step": 348
    },
    {
      "epoch": 0.12040710712437468,
      "grad_norm": 11.094878196716309,
      "learning_rate": 6.30830273062464e-05,
      "loss": 8.3762,
      "step": 349
    },
    {
      "epoch": 0.12075211316198034,
      "grad_norm": 12.04594612121582,
      "learning_rate": 6.234897333112594e-05,
      "loss": 8.091,
      "step": 350
    },
    {
      "epoch": 0.12109711919958599,
      "grad_norm": 11.683172225952148,
      "learning_rate": 6.161730353339302e-05,
      "loss": 9.0772,
      "step": 351
    },
    {
      "epoch": 0.12144212523719165,
      "grad_norm": 11.814870834350586,
      "learning_rate": 6.088806304588717e-05,
      "loss": 9.3354,
      "step": 352
    },
    {
      "epoch": 0.1217871312747973,
      "grad_norm": 11.268485069274902,
      "learning_rate": 6.0161296851596766e-05,
      "loss": 8.9482,
      "step": 353
    },
    {
      "epoch": 0.12213213731240297,
      "grad_norm": 11.454604148864746,
      "learning_rate": 5.943704978088402e-05,
      "loss": 8.9604,
      "step": 354
    },
    {
      "epoch": 0.12247714335000863,
      "grad_norm": 12.425684928894043,
      "learning_rate": 5.871536650871979e-05,
      "loss": 8.926,
      "step": 355
    },
    {
      "epoch": 0.12282214938761428,
      "grad_norm": 11.803454399108887,
      "learning_rate": 5.7996291551927666e-05,
      "loss": 9.0383,
      "step": 356
    },
    {
      "epoch": 0.12316715542521994,
      "grad_norm": 14.023934364318848,
      "learning_rate": 5.7279869266438234e-05,
      "loss": 9.2864,
      "step": 357
    },
    {
      "epoch": 0.12351216146282559,
      "grad_norm": 12.129143714904785,
      "learning_rate": 5.656614384455257e-05,
      "loss": 9.1773,
      "step": 358
    },
    {
      "epoch": 0.12385716750043126,
      "grad_norm": 11.97780990600586,
      "learning_rate": 5.585515931221677e-05,
      "loss": 8.5026,
      "step": 359
    },
    {
      "epoch": 0.12420217353803692,
      "grad_norm": 11.319969177246094,
      "learning_rate": 5.514695952630578e-05,
      "loss": 8.6805,
      "step": 360
    },
    {
      "epoch": 0.12454717957564257,
      "grad_norm": 10.946098327636719,
      "learning_rate": 5.444158817191832e-05,
      "loss": 9.0371,
      "step": 361
    },
    {
      "epoch": 0.12489218561324823,
      "grad_norm": 11.960458755493164,
      "learning_rate": 5.373908875968211e-05,
      "loss": 8.47,
      "step": 362
    },
    {
      "epoch": 0.1252371916508539,
      "grad_norm": 11.080780029296875,
      "learning_rate": 5.3039504623069965e-05,
      "loss": 9.7006,
      "step": 363
    },
    {
      "epoch": 0.12558219768845955,
      "grad_norm": 11.652853965759277,
      "learning_rate": 5.234287891572674e-05,
      "loss": 9.1047,
      "step": 364
    },
    {
      "epoch": 0.1259272037260652,
      "grad_norm": 11.393309593200684,
      "learning_rate": 5.164925460880758e-05,
      "loss": 9.0363,
      "step": 365
    },
    {
      "epoch": 0.12627220976367087,
      "grad_norm": 11.433116912841797,
      "learning_rate": 5.095867448832683e-05,
      "loss": 8.4939,
      "step": 366
    },
    {
      "epoch": 0.12661721580127652,
      "grad_norm": 12.424209594726562,
      "learning_rate": 5.027118115251938e-05,
      "loss": 9.122,
      "step": 367
    },
    {
      "epoch": 0.12696222183888217,
      "grad_norm": 11.743790626525879,
      "learning_rate": 4.95868170092125e-05,
      "loss": 9.4844,
      "step": 368
    },
    {
      "epoch": 0.12730722787648785,
      "grad_norm": 12.026018142700195,
      "learning_rate": 4.890562427321021e-05,
      "loss": 9.4922,
      "step": 369
    },
    {
      "epoch": 0.1276522339140935,
      "grad_norm": 11.183802604675293,
      "learning_rate": 4.822764496368917e-05,
      "loss": 9.2116,
      "step": 370
    },
    {
      "epoch": 0.12799723995169915,
      "grad_norm": 11.574292182922363,
      "learning_rate": 4.755292090160676e-05,
      "loss": 8.8551,
      "step": 371
    },
    {
      "epoch": 0.12834224598930483,
      "grad_norm": 12.3942232131958,
      "learning_rate": 4.6881493707121315e-05,
      "loss": 9.2175,
      "step": 372
    },
    {
      "epoch": 0.12868725202691048,
      "grad_norm": 12.747774124145508,
      "learning_rate": 4.621340479702503e-05,
      "loss": 8.9362,
      "step": 373
    },
    {
      "epoch": 0.12903225806451613,
      "grad_norm": 10.580770492553711,
      "learning_rate": 4.554869538218868e-05,
      "loss": 8.6681,
      "step": 374
    },
    {
      "epoch": 0.12937726410212178,
      "grad_norm": 12.67744255065918,
      "learning_rate": 4.48874064650202e-05,
      "loss": 8.6465,
      "step": 375
    },
    {
      "epoch": 0.12972227013972745,
      "grad_norm": 11.992030143737793,
      "learning_rate": 4.422957883693483e-05,
      "loss": 8.8035,
      "step": 376
    },
    {
      "epoch": 0.1300672761773331,
      "grad_norm": 10.928321838378906,
      "learning_rate": 4.357525307583933e-05,
      "loss": 9.1568,
      "step": 377
    },
    {
      "epoch": 0.13041228221493875,
      "grad_norm": 12.544729232788086,
      "learning_rate": 4.29244695436289e-05,
      "loss": 8.8095,
      "step": 378
    },
    {
      "epoch": 0.13075728825254443,
      "grad_norm": 12.932316780090332,
      "learning_rate": 4.227726838369711e-05,
      "loss": 8.9637,
      "step": 379
    },
    {
      "epoch": 0.13110229429015008,
      "grad_norm": 12.030780792236328,
      "learning_rate": 4.1633689518460225e-05,
      "loss": 8.7272,
      "step": 380
    },
    {
      "epoch": 0.13144730032775573,
      "grad_norm": 12.82052993774414,
      "learning_rate": 4.0993772646894116e-05,
      "loss": 9.4796,
      "step": 381
    },
    {
      "epoch": 0.1317923063653614,
      "grad_norm": 12.270413398742676,
      "learning_rate": 4.035755724208573e-05,
      "loss": 9.1355,
      "step": 382
    },
    {
      "epoch": 0.13213731240296706,
      "grad_norm": 12.860034942626953,
      "learning_rate": 3.972508254879805e-05,
      "loss": 8.9216,
      "step": 383
    },
    {
      "epoch": 0.1324823184405727,
      "grad_norm": 10.022688865661621,
      "learning_rate": 3.90963875810494e-05,
      "loss": 8.6796,
      "step": 384
    },
    {
      "epoch": 0.13282732447817835,
      "grad_norm": 11.842927932739258,
      "learning_rate": 3.847151111970676e-05,
      "loss": 9.3846,
      "step": 385
    },
    {
      "epoch": 0.13317233051578403,
      "grad_norm": 11.043813705444336,
      "learning_rate": 3.785049171009381e-05,
      "loss": 8.8771,
      "step": 386
    },
    {
      "epoch": 0.13351733655338968,
      "grad_norm": 11.62500286102295,
      "learning_rate": 3.723336765961285e-05,
      "loss": 9.1445,
      "step": 387
    },
    {
      "epoch": 0.13386234259099533,
      "grad_norm": 10.94227409362793,
      "learning_rate": 3.662017703538234e-05,
      "loss": 8.544,
      "step": 388
    },
    {
      "epoch": 0.134207348628601,
      "grad_norm": 11.737895965576172,
      "learning_rate": 3.601095766188833e-05,
      "loss": 9.2022,
      "step": 389
    },
    {
      "epoch": 0.13455235466620666,
      "grad_norm": 11.71877670288086,
      "learning_rate": 3.540574711865146e-05,
      "loss": 9.7183,
      "step": 390
    },
    {
      "epoch": 0.1348973607038123,
      "grad_norm": 12.432657241821289,
      "learning_rate": 3.4804582737908825e-05,
      "loss": 9.0206,
      "step": 391
    },
    {
      "epoch": 0.13524236674141799,
      "grad_norm": 11.108407020568848,
      "learning_rate": 3.420750160231118e-05,
      "loss": 8.5036,
      "step": 392
    },
    {
      "epoch": 0.13558737277902363,
      "grad_norm": 11.27615737915039,
      "learning_rate": 3.361454054263541e-05,
      "loss": 9.0201,
      "step": 393
    },
    {
      "epoch": 0.13593237881662928,
      "grad_norm": 9.975693702697754,
      "learning_rate": 3.302573613551292e-05,
      "loss": 8.3719,
      "step": 394
    },
    {
      "epoch": 0.13627738485423496,
      "grad_norm": 10.459684371948242,
      "learning_rate": 3.244112470117288e-05,
      "loss": 7.7744,
      "step": 395
    },
    {
      "epoch": 0.1366223908918406,
      "grad_norm": 10.368301391601562,
      "learning_rate": 3.186074230120244e-05,
      "loss": 8.8836,
      "step": 396
    },
    {
      "epoch": 0.13696739692944626,
      "grad_norm": 11.138334274291992,
      "learning_rate": 3.1284624736321846e-05,
      "loss": 8.1079,
      "step": 397
    },
    {
      "epoch": 0.1373124029670519,
      "grad_norm": 10.96679401397705,
      "learning_rate": 3.071280754417626e-05,
      "loss": 9.1686,
      "step": 398
    },
    {
      "epoch": 0.1376574090046576,
      "grad_norm": 11.052117347717285,
      "learning_rate": 3.0145325997143577e-05,
      "loss": 9.5247,
      "step": 399
    },
    {
      "epoch": 0.13800241504226324,
      "grad_norm": 10.489035606384277,
      "learning_rate": 2.9582215100158706e-05,
      "loss": 8.7892,
      "step": 400
    },
    {
      "epoch": 0.1383474210798689,
      "grad_norm": 13.531486511230469,
      "learning_rate": 2.902350958855426e-05,
      "loss": 8.2834,
      "step": 401
    },
    {
      "epoch": 0.13869242711747456,
      "grad_norm": 11.792000770568848,
      "learning_rate": 2.846924392591794e-05,
      "loss": 8.9256,
      "step": 402
    },
    {
      "epoch": 0.13903743315508021,
      "grad_norm": 10.499308586120605,
      "learning_rate": 2.791945230196663e-05,
      "loss": 8.2067,
      "step": 403
    },
    {
      "epoch": 0.13938243919268586,
      "grad_norm": 12.931747436523438,
      "learning_rate": 2.7374168630437456e-05,
      "loss": 7.8484,
      "step": 404
    },
    {
      "epoch": 0.13972744523029154,
      "grad_norm": 12.082032203674316,
      "learning_rate": 2.6833426546995782e-05,
      "loss": 9.9827,
      "step": 405
    },
    {
      "epoch": 0.1400724512678972,
      "grad_norm": 10.981305122375488,
      "learning_rate": 2.629725940716041e-05,
      "loss": 8.2251,
      "step": 406
    },
    {
      "epoch": 0.14041745730550284,
      "grad_norm": 14.595072746276855,
      "learning_rate": 2.57657002842462e-05,
      "loss": 9.9597,
      "step": 407
    },
    {
      "epoch": 0.14076246334310852,
      "grad_norm": 10.586888313293457,
      "learning_rate": 2.523878196732358e-05,
      "loss": 8.8847,
      "step": 408
    },
    {
      "epoch": 0.14110746938071417,
      "grad_norm": 11.972230911254883,
      "learning_rate": 2.4716536959196462e-05,
      "loss": 8.7841,
      "step": 409
    },
    {
      "epoch": 0.14145247541831982,
      "grad_norm": 10.82121753692627,
      "learning_rate": 2.4198997474396877e-05,
      "loss": 9.0321,
      "step": 410
    },
    {
      "epoch": 0.14179748145592547,
      "grad_norm": 10.548514366149902,
      "learning_rate": 2.3686195437198112e-05,
      "loss": 8.2817,
      "step": 411
    },
    {
      "epoch": 0.14214248749353114,
      "grad_norm": 12.447595596313477,
      "learning_rate": 2.31781624796453e-05,
      "loss": 9.384,
      "step": 412
    },
    {
      "epoch": 0.1424874935311368,
      "grad_norm": 10.921360969543457,
      "learning_rate": 2.2674929939604332e-05,
      "loss": 9.0153,
      "step": 413
    },
    {
      "epoch": 0.14283249956874244,
      "grad_norm": 11.576778411865234,
      "learning_rate": 2.217652885882869e-05,
      "loss": 9.6826,
      "step": 414
    },
    {
      "epoch": 0.14317750560634812,
      "grad_norm": 12.150237083435059,
      "learning_rate": 2.1682989981044783e-05,
      "loss": 9.3615,
      "step": 415
    },
    {
      "epoch": 0.14352251164395377,
      "grad_norm": 11.33244514465332,
      "learning_rate": 2.119434375005527e-05,
      "loss": 9.0407,
      "step": 416
    },
    {
      "epoch": 0.14386751768155942,
      "grad_norm": 10.824562072753906,
      "learning_rate": 2.071062030786149e-05,
      "loss": 9.4305,
      "step": 417
    },
    {
      "epoch": 0.1442125237191651,
      "grad_norm": 11.711275100708008,
      "learning_rate": 2.0231849492803852e-05,
      "loss": 9.5424,
      "step": 418
    },
    {
      "epoch": 0.14455752975677075,
      "grad_norm": 11.251872062683105,
      "learning_rate": 1.9758060837721467e-05,
      "loss": 9.9104,
      "step": 419
    },
    {
      "epoch": 0.1449025357943764,
      "grad_norm": 11.443036079406738,
      "learning_rate": 1.928928356813032e-05,
      "loss": 9.1346,
      "step": 420
    },
    {
      "epoch": 0.14524754183198205,
      "grad_norm": 13.695470809936523,
      "learning_rate": 1.882554660042052e-05,
      "loss": 9.6973,
      "step": 421
    },
    {
      "epoch": 0.14559254786958772,
      "grad_norm": 11.110825538635254,
      "learning_rate": 1.8366878540072614e-05,
      "loss": 9.876,
      "step": 422
    },
    {
      "epoch": 0.14593755390719337,
      "grad_norm": 10.91757583618164,
      "learning_rate": 1.7913307679893173e-05,
      "loss": 8.1501,
      "step": 423
    },
    {
      "epoch": 0.14628255994479902,
      "grad_norm": 11.609445571899414,
      "learning_rate": 1.7464861998269243e-05,
      "loss": 9.1033,
      "step": 424
    },
    {
      "epoch": 0.1466275659824047,
      "grad_norm": 13.72152328491211,
      "learning_rate": 1.702156915744292e-05,
      "loss": 9.9039,
      "step": 425
    },
    {
      "epoch": 0.14697257202001035,
      "grad_norm": 11.731778144836426,
      "learning_rate": 1.6583456501804725e-05,
      "loss": 9.6515,
      "step": 426
    },
    {
      "epoch": 0.147317578057616,
      "grad_norm": 11.61375617980957,
      "learning_rate": 1.6150551056206867e-05,
      "loss": 9.1904,
      "step": 427
    },
    {
      "epoch": 0.14766258409522168,
      "grad_norm": 10.265396118164062,
      "learning_rate": 1.57228795242965e-05,
      "loss": 8.8068,
      "step": 428
    },
    {
      "epoch": 0.14800759013282733,
      "grad_norm": 11.410909652709961,
      "learning_rate": 1.5300468286868137e-05,
      "loss": 8.8903,
      "step": 429
    },
    {
      "epoch": 0.14835259617043298,
      "grad_norm": 10.048901557922363,
      "learning_rate": 1.488334340023669e-05,
      "loss": 7.7431,
      "step": 430
    },
    {
      "epoch": 0.14869760220803865,
      "grad_norm": 11.77112102508545,
      "learning_rate": 1.4471530594629996e-05,
      "loss": 8.4086,
      "step": 431
    },
    {
      "epoch": 0.1490426082456443,
      "grad_norm": 10.68566608428955,
      "learning_rate": 1.4065055272601703e-05,
      "loss": 9.0967,
      "step": 432
    },
    {
      "epoch": 0.14938761428324995,
      "grad_norm": 10.029144287109375,
      "learning_rate": 1.3663942507464348e-05,
      "loss": 8.5118,
      "step": 433
    },
    {
      "epoch": 0.1497326203208556,
      "grad_norm": 10.815312385559082,
      "learning_rate": 1.3268217041742701e-05,
      "loss": 9.6862,
      "step": 434
    },
    {
      "epoch": 0.15007762635846128,
      "grad_norm": 10.653841972351074,
      "learning_rate": 1.2877903285647486e-05,
      "loss": 9.0271,
      "step": 435
    },
    {
      "epoch": 0.15042263239606693,
      "grad_norm": 11.70065689086914,
      "learning_rate": 1.2493025315569801e-05,
      "loss": 8.8587,
      "step": 436
    },
    {
      "epoch": 0.15076763843367258,
      "grad_norm": 11.202423095703125,
      "learning_rate": 1.2113606872595673e-05,
      "loss": 8.7797,
      "step": 437
    },
    {
      "epoch": 0.15111264447127826,
      "grad_norm": 10.97381591796875,
      "learning_rate": 1.173967136104196e-05,
      "loss": 9.584,
      "step": 438
    },
    {
      "epoch": 0.1514576505088839,
      "grad_norm": 10.520644187927246,
      "learning_rate": 1.1371241847012401e-05,
      "loss": 9.1783,
      "step": 439
    },
    {
      "epoch": 0.15180265654648956,
      "grad_norm": 13.997688293457031,
      "learning_rate": 1.1008341056974854e-05,
      "loss": 9.9617,
      "step": 440
    },
    {
      "epoch": 0.15214766258409523,
      "grad_norm": 11.452803611755371,
      "learning_rate": 1.0650991376359473e-05,
      "loss": 7.8551,
      "step": 441
    },
    {
      "epoch": 0.15249266862170088,
      "grad_norm": 11.610639572143555,
      "learning_rate": 1.029921484817783e-05,
      "loss": 8.7033,
      "step": 442
    },
    {
      "epoch": 0.15283767465930653,
      "grad_norm": 11.468731880187988,
      "learning_rate": 9.953033171663175e-06,
      "loss": 7.2685,
      "step": 443
    },
    {
      "epoch": 0.15318268069691218,
      "grad_norm": 11.875832557678223,
      "learning_rate": 9.612467700932045e-06,
      "loss": 8.854,
      "step": 444
    },
    {
      "epoch": 0.15352768673451786,
      "grad_norm": 11.493409156799316,
      "learning_rate": 9.277539443666783e-06,
      "loss": 9.3013,
      "step": 445
    },
    {
      "epoch": 0.1538726927721235,
      "grad_norm": 11.232786178588867,
      "learning_rate": 8.948269059820025e-06,
      "loss": 9.4602,
      "step": 446
    },
    {
      "epoch": 0.15421769880972916,
      "grad_norm": 10.974271774291992,
      "learning_rate": 8.624676860340025e-06,
      "loss": 9.8452,
      "step": 447
    },
    {
      "epoch": 0.15456270484733484,
      "grad_norm": 11.244220733642578,
      "learning_rate": 8.306782805917904e-06,
      "loss": 9.4318,
      "step": 448
    },
    {
      "epoch": 0.15490771088494049,
      "grad_norm": 10.4705228805542,
      "learning_rate": 7.994606505756355e-06,
      "loss": 8.8069,
      "step": 449
    },
    {
      "epoch": 0.15525271692254614,
      "grad_norm": 10.871196746826172,
      "learning_rate": 7.68816721636004e-06,
      "loss": 8.9918,
      "step": 450
    },
    {
      "epoch": 0.1555977229601518,
      "grad_norm": 10.249252319335938,
      "learning_rate": 7.3874838403478e-06,
      "loss": 9.3941,
      "step": 451
    },
    {
      "epoch": 0.15594272899775746,
      "grad_norm": 10.219954490661621,
      "learning_rate": 7.092574925286614e-06,
      "loss": 8.2221,
      "step": 452
    },
    {
      "epoch": 0.1562877350353631,
      "grad_norm": 12.183408737182617,
      "learning_rate": 6.803458662547507e-06,
      "loss": 9.0289,
      "step": 453
    },
    {
      "epoch": 0.1566327410729688,
      "grad_norm": 10.38846206665039,
      "learning_rate": 6.520152886183406e-06,
      "loss": 8.5604,
      "step": 454
    },
    {
      "epoch": 0.15697774711057444,
      "grad_norm": 11.054121017456055,
      "learning_rate": 6.242675071829111e-06,
      "loss": 8.0675,
      "step": 455
    },
    {
      "epoch": 0.1573227531481801,
      "grad_norm": 11.957045555114746,
      "learning_rate": 5.971042335623229e-06,
      "loss": 8.3669,
      "step": 456
    },
    {
      "epoch": 0.15766775918578574,
      "grad_norm": 11.593701362609863,
      "learning_rate": 5.705271433152458e-06,
      "loss": 9.2474,
      "step": 457
    },
    {
      "epoch": 0.15801276522339142,
      "grad_norm": 10.587969779968262,
      "learning_rate": 5.445378758417925e-06,
      "loss": 8.7036,
      "step": 458
    },
    {
      "epoch": 0.15835777126099707,
      "grad_norm": 11.409073829650879,
      "learning_rate": 5.191380342824035e-06,
      "loss": 8.8302,
      "step": 459
    },
    {
      "epoch": 0.15870277729860272,
      "grad_norm": 11.732986450195312,
      "learning_rate": 4.943291854189493e-06,
      "loss": 9.5684,
      "step": 460
    },
    {
      "epoch": 0.1590477833362084,
      "grad_norm": 12.807794570922852,
      "learning_rate": 4.701128595780878e-06,
      "loss": 9.3755,
      "step": 461
    },
    {
      "epoch": 0.15939278937381404,
      "grad_norm": 11.348825454711914,
      "learning_rate": 4.464905505368658e-06,
      "loss": 9.0037,
      "step": 462
    },
    {
      "epoch": 0.1597377954114197,
      "grad_norm": 11.585838317871094,
      "learning_rate": 4.23463715430577e-06,
      "loss": 8.6236,
      "step": 463
    },
    {
      "epoch": 0.16008280144902537,
      "grad_norm": 10.978170394897461,
      "learning_rate": 4.010337746628751e-06,
      "loss": 7.7244,
      "step": 464
    },
    {
      "epoch": 0.16042780748663102,
      "grad_norm": 11.80128002166748,
      "learning_rate": 3.792021118181636e-06,
      "loss": 9.2553,
      "step": 465
    },
    {
      "epoch": 0.16077281352423667,
      "grad_norm": 11.456130027770996,
      "learning_rate": 3.5797007357623945e-06,
      "loss": 8.8334,
      "step": 466
    },
    {
      "epoch": 0.16111781956184235,
      "grad_norm": 10.783517837524414,
      "learning_rate": 3.3733896962923658e-06,
      "loss": 8.4372,
      "step": 467
    },
    {
      "epoch": 0.161462825599448,
      "grad_norm": 10.7294282913208,
      "learning_rate": 3.1731007260082616e-06,
      "loss": 7.9268,
      "step": 468
    },
    {
      "epoch": 0.16180783163705365,
      "grad_norm": 11.667708396911621,
      "learning_rate": 2.9788461796772114e-06,
      "loss": 8.5356,
      "step": 469
    },
    {
      "epoch": 0.1621528376746593,
      "grad_norm": 12.107206344604492,
      "learning_rate": 2.790638039834668e-06,
      "loss": 9.2888,
      "step": 470
    },
    {
      "epoch": 0.16249784371226497,
      "grad_norm": 12.419697761535645,
      "learning_rate": 2.6084879160452166e-06,
      "loss": 9.1768,
      "step": 471
    },
    {
      "epoch": 0.16284284974987062,
      "grad_norm": 13.410768508911133,
      "learning_rate": 2.432407044186509e-06,
      "loss": 8.6753,
      "step": 472
    },
    {
      "epoch": 0.16318785578747627,
      "grad_norm": 12.53187370300293,
      "learning_rate": 2.26240628575615e-06,
      "loss": 9.184,
      "step": 473
    },
    {
      "epoch": 0.16353286182508195,
      "grad_norm": 10.773929595947266,
      "learning_rate": 2.098496127201648e-06,
      "loss": 8.5493,
      "step": 474
    },
    {
      "epoch": 0.1638778678626876,
      "grad_norm": 10.798660278320312,
      "learning_rate": 1.9406866792737267e-06,
      "loss": 8.8048,
      "step": 475
    },
    {
      "epoch": 0.16422287390029325,
      "grad_norm": 11.903676986694336,
      "learning_rate": 1.7889876764024505e-06,
      "loss": 9.1647,
      "step": 476
    },
    {
      "epoch": 0.16456787993789893,
      "grad_norm": 11.205957412719727,
      "learning_rate": 1.6434084760968697e-06,
      "loss": 8.4799,
      "step": 477
    },
    {
      "epoch": 0.16491288597550458,
      "grad_norm": 12.998312950134277,
      "learning_rate": 1.5039580583678393e-06,
      "loss": 8.6625,
      "step": 478
    },
    {
      "epoch": 0.16525789201311022,
      "grad_norm": 11.00870418548584,
      "learning_rate": 1.3706450251739613e-06,
      "loss": 8.5881,
      "step": 479
    },
    {
      "epoch": 0.16560289805071587,
      "grad_norm": 11.128656387329102,
      "learning_rate": 1.2434775998910964e-06,
      "loss": 9.1874,
      "step": 480
    },
    {
      "epoch": 0.16594790408832155,
      "grad_norm": 11.363465309143066,
      "learning_rate": 1.1224636268050439e-06,
      "loss": 8.7322,
      "step": 481
    },
    {
      "epoch": 0.1662929101259272,
      "grad_norm": 11.690573692321777,
      "learning_rate": 1.0076105706276888e-06,
      "loss": 8.5356,
      "step": 482
    },
    {
      "epoch": 0.16663791616353285,
      "grad_norm": 11.853788375854492,
      "learning_rate": 8.989255160365527e-07,
      "loss": 8.5468,
      "step": 483
    },
    {
      "epoch": 0.16698292220113853,
      "grad_norm": 10.81377124786377,
      "learning_rate": 7.964151672377458e-07,
      "loss": 8.6265,
      "step": 484
    },
    {
      "epoch": 0.16732792823874418,
      "grad_norm": 10.152823448181152,
      "learning_rate": 7.000858475524444e-07,
      "loss": 8.9867,
      "step": 485
    },
    {
      "epoch": 0.16767293427634983,
      "grad_norm": 11.468990325927734,
      "learning_rate": 6.099434990268609e-07,
      "loss": 8.6961,
      "step": 486
    },
    {
      "epoch": 0.1680179403139555,
      "grad_norm": 12.72700309753418,
      "learning_rate": 5.259936820656257e-07,
      "loss": 9.2915,
      "step": 487
    },
    {
      "epoch": 0.16836294635156115,
      "grad_norm": 13.002864837646484,
      "learning_rate": 4.482415750889204e-07,
      "loss": 9.7992,
      "step": 488
    },
    {
      "epoch": 0.1687079523891668,
      "grad_norm": 10.479470252990723,
      "learning_rate": 3.766919742129331e-07,
      "loss": 7.901,
      "step": 489
    },
    {
      "epoch": 0.16905295842677248,
      "grad_norm": 11.835942268371582,
      "learning_rate": 3.1134929295407564e-07,
      "loss": 8.9412,
      "step": 490
    },
    {
      "epoch": 0.16939796446437813,
      "grad_norm": 11.374271392822266,
      "learning_rate": 2.5221756195672563e-07,
      "loss": 9.096,
      "step": 491
    },
    {
      "epoch": 0.16974297050198378,
      "grad_norm": 12.500724792480469,
      "learning_rate": 1.9930042874457254e-07,
      "loss": 8.0155,
      "step": 492
    },
    {
      "epoch": 0.17008797653958943,
      "grad_norm": 11.962935447692871,
      "learning_rate": 1.5260115749566882e-07,
      "loss": 8.4747,
      "step": 493
    },
    {
      "epoch": 0.1704329825771951,
      "grad_norm": 12.082106590270996,
      "learning_rate": 1.1212262884103974e-07,
      "loss": 9.5219,
      "step": 494
    },
    {
      "epoch": 0.17077798861480076,
      "grad_norm": 11.523614883422852,
      "learning_rate": 7.7867339686987e-08,
      "loss": 8.6196,
      "step": 495
    },
    {
      "epoch": 0.1711229946524064,
      "grad_norm": 10.092193603515625,
      "learning_rate": 4.98374030611084e-08,
      "loss": 8.7468,
      "step": 496
    },
    {
      "epoch": 0.17146800069001208,
      "grad_norm": 12.635038375854492,
      "learning_rate": 2.8034547981943713e-08,
      "loss": 9.0217,
      "step": 497
    },
    {
      "epoch": 0.17181300672761773,
      "grad_norm": 10.622178077697754,
      "learning_rate": 1.246011935228064e-08,
      "loss": 8.9715,
      "step": 498
    },
    {
      "epoch": 0.17215801276522338,
      "grad_norm": 11.02623176574707,
      "learning_rate": 3.115077876243988e-09,
      "loss": 9.5219,
      "step": 499
    },
    {
      "epoch": 0.17250301880282906,
      "grad_norm": 11.338805198669434,
      "learning_rate": 0.0,
      "loss": 7.7523,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.45348666367017e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}