{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2181,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 130.6673912475082,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 9.4922,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 134.54634966851276,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 9.5859,
      "step": 2
    },
    {
      "epoch": 0.0,
      "grad_norm": 70.75092527033861,
      "learning_rate": 4.545454545454546e-05,
      "loss": 8.8281,
      "step": 3
    },
    {
      "epoch": 0.0,
      "grad_norm": 45.8287180014533,
      "learning_rate": 6.060606060606061e-05,
      "loss": 7.4062,
      "step": 4
    },
    {
      "epoch": 0.0,
      "grad_norm": 32.838332293773455,
      "learning_rate": 7.575757575757576e-05,
      "loss": 6.5352,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 15.371501279321766,
      "learning_rate": 9.090909090909092e-05,
      "loss": 5.957,
      "step": 6
    },
    {
      "epoch": 0.0,
      "grad_norm": 12.514989906442034,
      "learning_rate": 0.00010606060606060606,
      "loss": 5.7773,
      "step": 7
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.729155863116588,
      "learning_rate": 0.00012121212121212122,
      "loss": 5.668,
      "step": 8
    },
    {
      "epoch": 0.0,
      "grad_norm": 7.554667026547262,
      "learning_rate": 0.00013636363636363637,
      "loss": 5.457,
      "step": 9
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.136831709740585,
      "learning_rate": 0.00015151515151515152,
      "loss": 5.2344,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.3654015252579335,
      "learning_rate": 0.00016666666666666666,
      "loss": 5.1094,
      "step": 11
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.414037082079901,
      "learning_rate": 0.00018181818181818183,
      "loss": 4.9531,
      "step": 12
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.9131807418603035,
      "learning_rate": 0.00019696969696969695,
      "loss": 4.8242,
      "step": 13
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.2360579970592354,
      "learning_rate": 0.00021212121212121213,
      "loss": 4.8242,
      "step": 14
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.0580225446312927,
      "learning_rate": 0.00022727272727272727,
      "loss": 4.6953,
      "step": 15
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.73092678964882,
      "learning_rate": 0.00024242424242424245,
      "loss": 4.6094,
      "step": 16
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.7525209968933337,
      "learning_rate": 0.00025757575757575756,
      "loss": 4.5312,
      "step": 17
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.250059273054913,
      "learning_rate": 0.00027272727272727274,
      "loss": 4.5078,
      "step": 18
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.7403343453748756,
      "learning_rate": 0.0002878787878787879,
      "loss": 4.4453,
      "step": 19
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.954837856655733,
      "learning_rate": 0.00030303030303030303,
      "loss": 4.2773,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6349314932495371,
      "learning_rate": 0.0003181818181818182,
      "loss": 4.3242,
      "step": 21
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4774169089783695,
      "learning_rate": 0.0003333333333333333,
      "loss": 4.3086,
      "step": 22
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5553787222171778,
      "learning_rate": 0.0003484848484848485,
      "loss": 4.2051,
      "step": 23
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4022536604136704,
      "learning_rate": 0.00036363636363636367,
      "loss": 4.1641,
      "step": 24
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.2030742092392146,
      "learning_rate": 0.0003787878787878788,
      "loss": 4.2383,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9774235239474098,
      "learning_rate": 0.0003939393939393939,
      "loss": 4.0527,
      "step": 26
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9627503462825687,
      "learning_rate": 0.00040909090909090913,
      "loss": 4.002,
      "step": 27
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1624123088307292,
      "learning_rate": 0.00042424242424242425,
      "loss": 4.0957,
      "step": 28
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5071382171874308,
      "learning_rate": 0.0004393939393939394,
      "loss": 4.0859,
      "step": 29
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7892395554473574,
      "learning_rate": 0.00045454545454545455,
      "loss": 4.043,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9123039911787758,
      "learning_rate": 0.0004696969696969697,
      "loss": 3.9473,
      "step": 31
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.716783372410952,
      "learning_rate": 0.0004848484848484849,
      "loss": 3.9551,
      "step": 32
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8326054628167971,
      "learning_rate": 0.0005,
      "loss": 4.0234,
      "step": 33
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7052325820014936,
      "learning_rate": 0.0005151515151515151,
      "loss": 3.8516,
      "step": 34
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7148596222882666,
      "learning_rate": 0.0005303030303030302,
      "loss": 3.9414,
      "step": 35
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7183457679094106,
      "learning_rate": 0.0005454545454545455,
      "loss": 3.8984,
      "step": 36
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7408409642024854,
      "learning_rate": 0.0005606060606060606,
      "loss": 3.8398,
      "step": 37
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8874047782274374,
      "learning_rate": 0.0005757575757575758,
      "loss": 3.8945,
      "step": 38
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6398638654276685,
      "learning_rate": 0.0005909090909090909,
      "loss": 3.7344,
      "step": 39
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6810368413911023,
      "learning_rate": 0.0006060606060606061,
      "loss": 3.7402,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8313546179684439,
      "learning_rate": 0.0006212121212121212,
      "loss": 3.6992,
      "step": 41
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8416468181514648,
      "learning_rate": 0.0006363636363636364,
      "loss": 3.7598,
      "step": 42
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9177053598221669,
      "learning_rate": 0.0006515151515151515,
      "loss": 3.623,
      "step": 43
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7759520539432314,
      "learning_rate": 0.0006666666666666666,
      "loss": 3.6738,
      "step": 44
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.62938944077773,
      "learning_rate": 0.0006818181818181818,
      "loss": 3.6191,
      "step": 45
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6392390678259651,
      "learning_rate": 0.000696969696969697,
      "loss": 3.5039,
      "step": 46
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.5299802746197122,
      "learning_rate": 0.0007121212121212122,
      "loss": 3.502,
      "step": 47
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.5550632052325813,
      "learning_rate": 0.0007272727272727273,
      "loss": 3.5254,
      "step": 48
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.48045958853034776,
      "learning_rate": 0.0007424242424242425,
      "loss": 3.4707,
      "step": 49
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.48146859374710244,
      "learning_rate": 0.0007575757575757576,
      "loss": 3.4082,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.4592824205529459,
      "learning_rate": 0.0007727272727272727,
      "loss": 3.3691,
      "step": 51
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.4050043748098675,
      "learning_rate": 0.0007878787878787878,
      "loss": 3.543,
      "step": 52
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.37310094647856834,
      "learning_rate": 0.000803030303030303,
      "loss": 3.4102,
      "step": 53
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.38659977575716475,
      "learning_rate": 0.0008181818181818183,
      "loss": 3.4922,
      "step": 54
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.42387620637138484,
      "learning_rate": 0.0008333333333333334,
      "loss": 3.4277,
      "step": 55
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.43671973058260677,
      "learning_rate": 0.0008484848484848485,
      "loss": 3.3984,
      "step": 56
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3972045145106085,
      "learning_rate": 0.0008636363636363636,
      "loss": 3.418,
      "step": 57
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3866039492214528,
      "learning_rate": 0.0008787878787878789,
      "loss": 3.4648,
      "step": 58
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3291986992365848,
      "learning_rate": 0.000893939393939394,
      "loss": 3.2812,
      "step": 59
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3471965940208802,
      "learning_rate": 0.0009090909090909091,
      "loss": 3.3223,
      "step": 60
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3645973832887824,
      "learning_rate": 0.0009242424242424242,
      "loss": 3.4863,
      "step": 61
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3069535373287153,
      "learning_rate": 0.0009393939393939394,
      "loss": 3.3809,
      "step": 62
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3680117281724585,
      "learning_rate": 0.0009545454545454546,
      "loss": 3.4336,
      "step": 63
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.2875454973738176,
      "learning_rate": 0.0009696969696969698,
      "loss": 3.3555,
      "step": 64
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.24852499790032084,
      "learning_rate": 0.000984848484848485,
      "loss": 3.2988,
      "step": 65
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.2758556144946782,
      "learning_rate": 0.001,
      "loss": 3.293,
      "step": 66
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.32958955557245445,
      "learning_rate": 0.0009999994484067654,
      "loss": 3.3281,
      "step": 67
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.30098227559570195,
      "learning_rate": 0.0009999977936282788,
      "loss": 3.3574,
      "step": 68
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.29823132191448415,
      "learning_rate": 0.0009999950356681913,
      "loss": 3.4004,
      "step": 69
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.29997551970294084,
      "learning_rate": 0.0009999911745325876,
      "loss": 3.3516,
      "step": 70
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.30006935820373487,
      "learning_rate": 0.0009999862102299873,
      "loss": 3.3203,
      "step": 71
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.28327915130379644,
      "learning_rate": 0.0009999801427713433,
      "loss": 3.3086,
      "step": 72
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.24415879671448645,
      "learning_rate": 0.0009999729721700424,
      "loss": 3.2617,
      "step": 73
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.25707515693757343,
      "learning_rate": 0.000999964698441906,
      "loss": 3.2852,
      "step": 74
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.28060645018034047,
      "learning_rate": 0.0009999553216051892,
      "loss": 3.248,
      "step": 75
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.2543877400623929,
      "learning_rate": 0.00099994484168058,
      "loss": 3.3574,
      "step": 76
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.2682483162123672,
      "learning_rate": 0.0009999332586912019,
      "loss": 3.4023,
      "step": 77
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.32089461585633444,
      "learning_rate": 0.0009999205726626108,
      "loss": 3.2891,
      "step": 78
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.2567902845497091,
      "learning_rate": 0.000999906783622797,
      "loss": 3.2832,
      "step": 79
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.278411138661758,
      "learning_rate": 0.0009998918916021842,
      "loss": 3.2969,
      "step": 80
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.2800065872210731,
      "learning_rate": 0.0009998758966336297,
      "loss": 3.3047,
      "step": 81
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3048402617876785,
      "learning_rate": 0.0009998587987524242,
      "loss": 3.2969,
      "step": 82
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.28556478543122765,
      "learning_rate": 0.0009998405979962926,
      "loss": 3.2441,
      "step": 83
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.2994684141412536,
      "learning_rate": 0.000999821294405392,
      "loss": 3.3867,
      "step": 84
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.28529445564586065,
      "learning_rate": 0.0009998008880223134,
      "loss": 3.2539,
      "step": 85
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.23972264856650938,
      "learning_rate": 0.000999779378892081,
      "loss": 3.3418,
      "step": 86
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.27452511583633077,
      "learning_rate": 0.0009997567670621522,
      "loss": 3.2812,
      "step": 87
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.28553646343001515,
      "learning_rate": 0.0009997330525824165,
      "loss": 3.2812,
      "step": 88
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3056105652443864,
      "learning_rate": 0.0009997082355051976,
      "loss": 3.3047,
      "step": 89
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.24632654504196713,
      "learning_rate": 0.000999682315885251,
      "loss": 3.2266,
      "step": 90
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.29824465959288454,
      "learning_rate": 0.0009996552937797645,
      "loss": 3.3164,
      "step": 91
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.254819519816752,
      "learning_rate": 0.0009996271692483596,
      "loss": 3.2695,
      "step": 92
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.265105057593198,
      "learning_rate": 0.0009995979423530893,
      "loss": 3.2578,
      "step": 93
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.24685761240600235,
      "learning_rate": 0.000999567613158439,
      "loss": 3.1602,
      "step": 94
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.31855436487530725,
      "learning_rate": 0.0009995361817313263,
      "loss": 3.2305,
      "step": 95
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3300120784572777,
      "learning_rate": 0.0009995036481411004,
      "loss": 3.2793,
      "step": 96
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.2885552950730676,
      "learning_rate": 0.0009994700124595429,
      "loss": 3.2402,
      "step": 97
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.2982277283693368,
      "learning_rate": 0.0009994352747608663,
      "loss": 3.2266,
      "step": 98
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.30071697612379733,
      "learning_rate": 0.0009993994351217151,
      "loss": 3.25,
      "step": 99
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.28838113869089127,
      "learning_rate": 0.000999362493621165,
      "loss": 3.1816,
      "step": 100
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.30811398503557375,
      "learning_rate": 0.0009993244503407226,
      "loss": 3.2441,
      "step": 101
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.31897006973953634,
      "learning_rate": 0.0009992853053643258,
      "loss": 3.2148,
      "step": 102
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2852505840455887,
      "learning_rate": 0.0009992450587783426,
      "loss": 3.2207,
      "step": 103
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3094337290625498,
      "learning_rate": 0.000999203710671572,
      "loss": 3.1699,
      "step": 104
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3504615609721934,
      "learning_rate": 0.0009991612611352438,
      "loss": 3.1875,
      "step": 105
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.30544464664733656,
      "learning_rate": 0.0009991177102630173,
      "loss": 3.3008,
      "step": 106
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.30834227988098617,
      "learning_rate": 0.0009990730581509817,
      "loss": 3.127,
      "step": 107
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3066405437910335,
      "learning_rate": 0.0009990273048976566,
      "loss": 3.2266,
      "step": 108
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.33067110597283705,
      "learning_rate": 0.0009989804506039905,
      "loss": 3.1523,
      "step": 109
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3013788961808186,
      "learning_rate": 0.0009989324953733614,
      "loss": 3.3066,
      "step": 110
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2941271723477759,
      "learning_rate": 0.0009988834393115766,
      "loss": 3.2246,
      "step": 111
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3666862777659565,
      "learning_rate": 0.000998833282526872,
      "loss": 3.2402,
      "step": 112
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3108840790558172,
      "learning_rate": 0.0009987820251299122,
      "loss": 3.2695,
      "step": 113
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.31566791464808863,
      "learning_rate": 0.00099872966723379,
      "loss": 3.2109,
      "step": 114
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.33977367595270186,
      "learning_rate": 0.0009986762089540266,
      "loss": 3.2383,
      "step": 115
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3275180521716002,
      "learning_rate": 0.0009986216504085709,
      "loss": 3.2695,
      "step": 116
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.32116102197696306,
      "learning_rate": 0.0009985659917177991,
      "loss": 3.2227,
      "step": 117
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3579200278892064,
      "learning_rate": 0.0009985092330045155,
      "loss": 3.2344,
      "step": 118
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3222443598390205,
      "learning_rate": 0.0009984513743939508,
      "loss": 3.1172,
      "step": 119
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3671154398007888,
      "learning_rate": 0.0009983924160137626,
      "loss": 3.2051,
      "step": 120
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.33460343549017446,
      "learning_rate": 0.000998332357994035,
      "loss": 3.2422,
      "step": 121
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.40415311246118213,
      "learning_rate": 0.0009982712004672786,
      "loss": 3.2578,
      "step": 122
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.378426424141603,
      "learning_rate": 0.0009982089435684295,
      "loss": 3.2578,
      "step": 123
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3465912481745706,
      "learning_rate": 0.0009981455874348499,
      "loss": 3.1406,
      "step": 124
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3519444585540837,
      "learning_rate": 0.0009980811322063269,
      "loss": 3.2285,
      "step": 125
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3389580400789743,
      "learning_rate": 0.0009980155780250728,
      "loss": 3.2695,
      "step": 126
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3403476317989102,
      "learning_rate": 0.0009979489250357243,
      "loss": 3.2559,
      "step": 127
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3743267851159314,
      "learning_rate": 0.0009978811733853431,
      "loss": 3.2383,
      "step": 128
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3948602803984734,
      "learning_rate": 0.0009978123232234147,
      "loss": 3.1289,
      "step": 129
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.477802326300184,
      "learning_rate": 0.000997742374701848,
      "loss": 3.3359,
      "step": 130
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4185243541839902,
      "learning_rate": 0.0009976713279749754,
      "loss": 3.1914,
      "step": 131
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3244838234799754,
      "learning_rate": 0.0009975991831995528,
      "loss": 3.2324,
      "step": 132
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4693693945618756,
      "learning_rate": 0.0009975259405347581,
      "loss": 3.2617,
      "step": 133
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.36617875994937044,
      "learning_rate": 0.0009974516001421926,
      "loss": 3.1934,
      "step": 134
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3851167616939522,
      "learning_rate": 0.000997376162185878,
      "loss": 3.1797,
      "step": 135
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4338895439137645,
      "learning_rate": 0.0009972996268322594,
      "loss": 3.2578,
      "step": 136
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3949760366391964,
      "learning_rate": 0.0009972219942502017,
      "loss": 3.1914,
      "step": 137
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.41262887758641187,
      "learning_rate": 0.0009971432646109918,
      "loss": 3.1641,
      "step": 138
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3756736614013557,
      "learning_rate": 0.0009970634380883365,
      "loss": 3.1895,
      "step": 139
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4059667279807796,
      "learning_rate": 0.0009969825148583627,
      "loss": 3.2129,
      "step": 140
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4169914241899365,
      "learning_rate": 0.0009969004950996173,
      "loss": 3.3594,
      "step": 141
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4659200136689364,
      "learning_rate": 0.0009968173789930668,
      "loss": 3.2422,
      "step": 142
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4425567962515051,
      "learning_rate": 0.0009967331667220958,
      "loss": 3.209,
      "step": 143
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.44807700072424106,
      "learning_rate": 0.0009966478584725086,
      "loss": 3.2207,
      "step": 144
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.49536132850689274,
      "learning_rate": 0.0009965614544325263,
      "loss": 3.2246,
      "step": 145
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5430608251648995,
      "learning_rate": 0.000996473954792789,
      "loss": 3.1426,
      "step": 146
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.46668415979452554,
      "learning_rate": 0.0009963853597463532,
      "loss": 3.1582,
      "step": 147
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.40296724270981343,
      "learning_rate": 0.000996295669488693,
      "loss": 3.1211,
      "step": 148
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4361616051714609,
      "learning_rate": 0.0009962048842176979,
      "loss": 3.3184,
      "step": 149
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4075778740568562,
      "learning_rate": 0.0009961130041336748,
      "loss": 3.1582,
      "step": 150
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.49066147353163797,
      "learning_rate": 0.0009960200294393449,
      "loss": 3.1699,
      "step": 151
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4035771299197594,
      "learning_rate": 0.0009959259603398453,
      "loss": 3.207,
      "step": 152
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.3378459300807858,
      "learning_rate": 0.0009958307970427275,
      "loss": 3.2285,
      "step": 153
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4086864633607365,
      "learning_rate": 0.0009957345397579572,
      "loss": 3.248,
      "step": 154
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4141574920894895,
      "learning_rate": 0.0009956371886979138,
      "loss": 3.1523,
      "step": 155
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4256719786310397,
      "learning_rate": 0.00099553874407739,
      "loss": 3.1035,
      "step": 156
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.42154861993070003,
      "learning_rate": 0.0009954392061135916,
      "loss": 3.2363,
      "step": 157
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5177391063370709,
      "learning_rate": 0.0009953385750261364,
      "loss": 3.3203,
      "step": 158
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.42413272477131575,
      "learning_rate": 0.0009952368510370538,
      "loss": 3.1465,
      "step": 159
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.42741818912961904,
      "learning_rate": 0.0009951340343707852,
      "loss": 3.1484,
      "step": 160
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.46844796836680097,
      "learning_rate": 0.0009950301252541823,
      "loss": 3.1621,
      "step": 161
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5432069897331167,
      "learning_rate": 0.0009949251239165075,
      "loss": 3.1934,
      "step": 162
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4433243595809475,
      "learning_rate": 0.000994819030589433,
      "loss": 3.2441,
      "step": 163
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.46442240558242925,
      "learning_rate": 0.00099471184550704,
      "loss": 3.0859,
      "step": 164
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.44362209636880023,
      "learning_rate": 0.0009946035689058189,
      "loss": 3.1914,
      "step": 165
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3929409968411158,
      "learning_rate": 0.0009944942010246681,
      "loss": 3.1152,
      "step": 166
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.38690574116620396,
      "learning_rate": 0.0009943837421048942,
      "loss": 3.0859,
      "step": 167
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.46247613538293986,
      "learning_rate": 0.0009942721923902106,
      "loss": 3.1406,
      "step": 168
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4833631894296634,
      "learning_rate": 0.0009941595521267377,
      "loss": 3.1309,
      "step": 169
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3553583797986057,
      "learning_rate": 0.0009940458215630017,
      "loss": 3.1387,
      "step": 170
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3448909906308115,
      "learning_rate": 0.0009939310009499348,
      "loss": 3.1289,
      "step": 171
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3981193694506046,
      "learning_rate": 0.000993815090540874,
      "loss": 3.1484,
      "step": 172
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3750474696406515,
      "learning_rate": 0.000993698090591561,
      "loss": 3.1602,
      "step": 173
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.38595572316882176,
      "learning_rate": 0.0009935800013601416,
      "loss": 3.1445,
      "step": 174
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.40690037265563644,
      "learning_rate": 0.000993460823107164,
      "loss": 3.0586,
      "step": 175
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4720134105377351,
      "learning_rate": 0.0009933405560955803,
      "loss": 3.2129,
      "step": 176
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3597845184975643,
      "learning_rate": 0.0009932192005907446,
      "loss": 3.1465,
      "step": 177
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.378027139744156,
      "learning_rate": 0.0009930967568604118,
      "loss": 3.168,
      "step": 178
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.46799699857039784,
      "learning_rate": 0.000992973225174739,
      "loss": 3.1406,
      "step": 179
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4874504054301936,
      "learning_rate": 0.0009928486058062827,
      "loss": 3.1133,
      "step": 180
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3639696939229084,
      "learning_rate": 0.0009927228990299999,
      "loss": 3.1445,
      "step": 181
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4840942374431747,
      "learning_rate": 0.0009925961051232468,
      "loss": 3.0918,
      "step": 182
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5002798024360261,
      "learning_rate": 0.000992468224365778,
      "loss": 3.2051,
      "step": 183
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4163394051938708,
      "learning_rate": 0.000992339257039746,
      "loss": 3.1367,
      "step": 184
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.536518648413036,
      "learning_rate": 0.0009922092034297006,
      "loss": 3.1895,
      "step": 185
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5182302191381075,
      "learning_rate": 0.0009920780638225891,
      "loss": 3.0859,
      "step": 186
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.44203587212572937,
      "learning_rate": 0.0009919458385077538,
      "loss": 3.0664,
      "step": 187
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5183300810939687,
      "learning_rate": 0.0009918125277769336,
      "loss": 3.1133,
      "step": 188
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4719025861457211,
      "learning_rate": 0.0009916781319242614,
      "loss": 3.1328,
      "step": 189
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5162362886892993,
      "learning_rate": 0.0009915426512462646,
      "loss": 3.1055,
      "step": 190
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.443070929111309,
      "learning_rate": 0.0009914060860418644,
      "loss": 3.0391,
      "step": 191
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.41055240636734797,
      "learning_rate": 0.000991268436612374,
      "loss": 3.127,
      "step": 192
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5240295948680025,
      "learning_rate": 0.0009911297032614997,
      "loss": 3.1289,
      "step": 193
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.39407029726131976,
      "learning_rate": 0.000990989886295339,
      "loss": 3.1074,
      "step": 194
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.40897947196483103,
      "learning_rate": 0.0009908489860223804,
      "loss": 3.2324,
      "step": 195
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4551149509464412,
      "learning_rate": 0.000990707002753502,
      "loss": 3.1758,
      "step": 196
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4470627343548775,
      "learning_rate": 0.0009905639368019724,
      "loss": 3.1426,
      "step": 197
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.46109704603441126,
      "learning_rate": 0.0009904197884834482,
      "loss": 3.1016,
      "step": 198
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.42737322349816875,
      "learning_rate": 0.0009902745581159742,
      "loss": 3.0957,
      "step": 199
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4611923685158927,
      "learning_rate": 0.0009901282460199829,
      "loss": 3.2227,
      "step": 200
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5064632783447336,
      "learning_rate": 0.0009899808525182935,
      "loss": 3.2539,
      "step": 201
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.491401832984581,
      "learning_rate": 0.0009898323779361107,
      "loss": 3.1406,
      "step": 202
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.3978849593982214,
      "learning_rate": 0.000989682822601025,
      "loss": 3.1367,
      "step": 203
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.453293635329491,
      "learning_rate": 0.0009895321868430113,
      "loss": 3.1465,
      "step": 204
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5567913953135193,
      "learning_rate": 0.0009893804709944281,
      "loss": 3.2168,
      "step": 205
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.36934388798099055,
      "learning_rate": 0.0009892276753900174,
      "loss": 3.1562,
      "step": 206
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.44068830643325835,
      "learning_rate": 0.0009890738003669028,
      "loss": 3.1836,
      "step": 207
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5402900330792273,
      "learning_rate": 0.0009889188462645904,
      "loss": 3.1484,
      "step": 208
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5173515067186724,
      "learning_rate": 0.0009887628134249667,
      "loss": 3.0273,
      "step": 209
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.45448940466985077,
      "learning_rate": 0.0009886057021922983,
      "loss": 3.1309,
      "step": 210
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.43098077468561125,
      "learning_rate": 0.0009884475129132311,
      "loss": 3.1406,
      "step": 211
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4739137511003327,
      "learning_rate": 0.0009882882459367897,
      "loss": 3.1289,
      "step": 212
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4198572085456846,
      "learning_rate": 0.0009881279016143766,
      "loss": 3.1562,
      "step": 213
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5527214196575254,
      "learning_rate": 0.0009879664802997707,
      "loss": 3.123,
      "step": 214
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4461997726409922,
      "learning_rate": 0.000987803982349128,
      "loss": 3.0879,
      "step": 215
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4114908961375025,
      "learning_rate": 0.0009876404081209796,
      "loss": 3.0938,
      "step": 216
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5090883264037154,
      "learning_rate": 0.000987475757976231,
      "loss": 3.1309,
      "step": 217
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4832699974580469,
      "learning_rate": 0.000987310032278162,
      "loss": 3.0859,
      "step": 218
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.45308365731505645,
      "learning_rate": 0.0009871432313924254,
      "loss": 3.1738,
      "step": 219
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4575120366360754,
      "learning_rate": 0.000986975355687046,
      "loss": 3.1621,
      "step": 220
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.44581544074850504,
      "learning_rate": 0.0009868064055324204,
      "loss": 3.2129,
      "step": 221
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.45062690025815266,
      "learning_rate": 0.0009866363813013153,
      "loss": 3.1328,
      "step": 222
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4164668718638638,
      "learning_rate": 0.0009864652833688676,
      "loss": 3.1172,
      "step": 223
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.41988697946626796,
      "learning_rate": 0.0009862931121125836,
      "loss": 3.0723,
      "step": 224
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.406782903790899,
      "learning_rate": 0.000986119867912337,
      "loss": 3.0762,
      "step": 225
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.3875244543013597,
      "learning_rate": 0.000985945551150369,
      "loss": 3.1211,
      "step": 226
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4564051614400283,
      "learning_rate": 0.0009857701622112876,
      "loss": 3.1113,
      "step": 227
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.41971381722960416,
      "learning_rate": 0.000985593701482066,
      "loss": 3.1172,
      "step": 228
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.43265010777306967,
      "learning_rate": 0.0009854161693520424,
      "loss": 3.1328,
      "step": 229
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5045538861566129,
      "learning_rate": 0.0009852375662129194,
      "loss": 3.1875,
      "step": 230
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4295117303883956,
      "learning_rate": 0.0009850578924587613,
      "loss": 3.0996,
      "step": 231
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.43905644255399445,
      "learning_rate": 0.000984877148485996,
      "loss": 3.1289,
      "step": 232
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.42251980596494393,
      "learning_rate": 0.000984695334693412,
      "loss": 3.0938,
      "step": 233
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6104996954686958,
      "learning_rate": 0.000984512451482158,
      "loss": 3.1113,
      "step": 234
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.46225237812576064,
      "learning_rate": 0.0009843284992557431,
      "loss": 3.0605,
      "step": 235
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.520619327079272,
      "learning_rate": 0.000984143478420034,
      "loss": 3.1152,
      "step": 236
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4382662726022456,
      "learning_rate": 0.0009839573893832563,
      "loss": 3.0508,
      "step": 237
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.47124507748473143,
      "learning_rate": 0.000983770232555991,
      "loss": 3.0664,
      "step": 238
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4881718969530819,
      "learning_rate": 0.0009835820083511765,
      "loss": 3.0469,
      "step": 239
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.49153127209333136,
      "learning_rate": 0.0009833927171841055,
      "loss": 3.1328,
      "step": 240
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4237731035946726,
      "learning_rate": 0.0009832023594724246,
      "loss": 3.0918,
      "step": 241
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5217807598692051,
      "learning_rate": 0.0009830109356361344,
      "loss": 3.1191,
      "step": 242
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.48030191578585496,
      "learning_rate": 0.0009828184460975867,
      "loss": 3.1523,
      "step": 243
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.49482688823215715,
      "learning_rate": 0.0009826248912814855,
      "loss": 3.1055,
      "step": 244
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4466427155282997,
      "learning_rate": 0.0009824302716148847,
      "loss": 3.0801,
      "step": 245
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4973610548096533,
      "learning_rate": 0.0009822345875271884,
      "loss": 3.0957,
      "step": 246
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.43849518676383825,
      "learning_rate": 0.0009820378394501481,
      "loss": 3.1016,
      "step": 247
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.455105646564555,
      "learning_rate": 0.0009818400278178636,
      "loss": 3.0938,
      "step": 248
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.47797815180088643,
      "learning_rate": 0.0009816411530667814,
      "loss": 3.1465,
      "step": 249
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5446688179975222,
      "learning_rate": 0.000981441215635693,
      "loss": 3.1484,
      "step": 250
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5807668658263057,
      "learning_rate": 0.0009812402159657353,
      "loss": 3.2109,
      "step": 251
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.43344480318542294,
      "learning_rate": 0.000981038154500388,
      "loss": 3.1348,
      "step": 252
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5423304700110778,
      "learning_rate": 0.0009808350316854746,
      "loss": 3.1426,
      "step": 253
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5688400375486342,
      "learning_rate": 0.0009806308479691594,
      "loss": 3.0664,
      "step": 254
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5025398086902951,
      "learning_rate": 0.0009804256038019482,
      "loss": 3.1895,
      "step": 255
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.512574194611775,
      "learning_rate": 0.0009802192996366857,
      "loss": 3.0879,
      "step": 256
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.42040604726500763,
      "learning_rate": 0.0009800119359285563,
      "loss": 3.0801,
      "step": 257
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5028432904378699,
      "learning_rate": 0.0009798035131350813,
      "loss": 3.1875,
      "step": 258
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.43759310079798647,
      "learning_rate": 0.0009795940317161194,
      "loss": 3.1152,
      "step": 259
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.39942622574284586,
      "learning_rate": 0.0009793834921338646,
      "loss": 3.1914,
      "step": 260
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4313625675097166,
      "learning_rate": 0.0009791718948528457,
      "loss": 3.0801,
      "step": 261
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.49915342844749416,
      "learning_rate": 0.0009789592403399252,
      "loss": 3.0664,
      "step": 262
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4941994500755652,
      "learning_rate": 0.0009787455290642985,
      "loss": 3.2207,
      "step": 263
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.47020094876120216,
      "learning_rate": 0.000978530761497492,
      "loss": 3.1191,
      "step": 264
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4308752015555103,
      "learning_rate": 0.0009783149381133633,
      "loss": 3.1191,
      "step": 265
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4492218802857437,
      "learning_rate": 0.0009780980593880992,
      "loss": 3.1914,
      "step": 266
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4048772217628048,
      "learning_rate": 0.0009778801258002153,
      "loss": 3.1348,
      "step": 267
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.45718340553493436,
      "learning_rate": 0.000977661137830554,
      "loss": 3.0547,
      "step": 268
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5416924141090603,
      "learning_rate": 0.0009774410959622845,
      "loss": 3.1875,
      "step": 269
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5025625236690754,
      "learning_rate": 0.000977220000680901,
      "loss": 3.1777,
      "step": 270
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4481205685654053,
      "learning_rate": 0.000976997852474223,
      "loss": 3.1094,
      "step": 271
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.48913690115210856,
      "learning_rate": 0.0009767746518323914,
      "loss": 3.0664,
      "step": 272
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4183993959010507,
      "learning_rate": 0.0009765503992478704,
      "loss": 3.0547,
      "step": 273
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.38396505466503583,
      "learning_rate": 0.0009763250952154449,
      "loss": 3.0508,
      "step": 274
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4085021372600433,
      "learning_rate": 0.0009760987402322195,
      "loss": 2.9355,
      "step": 275
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.447310106512099,
      "learning_rate": 0.0009758713347976178,
      "loss": 3.0527,
      "step": 276
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.503367524129868,
      "learning_rate": 0.000975642879413381,
      "loss": 3.0352,
      "step": 277
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4822801807375108,
      "learning_rate": 0.0009754133745835665,
      "loss": 3.0996,
      "step": 278
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.43289339430340085,
      "learning_rate": 0.0009751828208145482,
      "loss": 3.1328,
      "step": 279
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5266235495881424,
      "learning_rate": 0.0009749512186150131,
      "loss": 3.1055,
      "step": 280
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.48979473303323556,
      "learning_rate": 0.0009747185684959625,
      "loss": 3.1406,
      "step": 281
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5115094035338374,
      "learning_rate": 0.000974484870970709,
      "loss": 3.1387,
      "step": 282
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5184612675780877,
      "learning_rate": 0.0009742501265548767,
      "loss": 3.0605,
      "step": 283
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.46085244759553234,
      "learning_rate": 0.0009740143357663993,
      "loss": 3.0352,
      "step": 284
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4387029986790845,
      "learning_rate": 0.000973777499125519,
      "loss": 3.209,
      "step": 285
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.49762445211441747,
      "learning_rate": 0.0009735396171547859,
      "loss": 3.1133,
      "step": 286
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4642844093743038,
      "learning_rate": 0.0009733006903790564,
      "loss": 3.0723,
      "step": 287
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4796883976054307,
      "learning_rate": 0.0009730607193254922,
      "loss": 3.084,
      "step": 288
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.45979416335057427,
      "learning_rate": 0.0009728197045235585,
      "loss": 3.1211,
      "step": 289
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4782616815163618,
      "learning_rate": 0.0009725776465050242,
      "loss": 3.1152,
      "step": 290
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5177364092714997,
      "learning_rate": 0.0009723345458039594,
      "loss": 3.0957,
      "step": 291
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.45390500017556673,
      "learning_rate": 0.000972090402956735,
      "loss": 3.0957,
      "step": 292
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5068376779468969,
      "learning_rate": 0.0009718452185020212,
      "loss": 2.998,
      "step": 293
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.45079139468161267,
      "learning_rate": 0.0009715989929807862,
      "loss": 3.0215,
      "step": 294
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5112996672226469,
      "learning_rate": 0.0009713517269362955,
      "loss": 2.9941,
      "step": 295
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5114556817761038,
      "learning_rate": 0.0009711034209141101,
      "loss": 3.0371,
      "step": 296
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5010614309022469,
      "learning_rate": 0.0009708540754620856,
      "loss": 3.0918,
      "step": 297
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4417990285590517,
      "learning_rate": 0.0009706036911303713,
      "loss": 3.123,
      "step": 298
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5434511325538337,
      "learning_rate": 0.0009703522684714083,
      "loss": 3.0371,
      "step": 299
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.43826392043631934,
      "learning_rate": 0.0009700998080399286,
      "loss": 3.0742,
      "step": 300
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5297666100088221,
      "learning_rate": 0.0009698463103929542,
      "loss": 3.0508,
      "step": 301
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4034288651363041,
      "learning_rate": 0.0009695917760897954,
      "loss": 3.0605,
      "step": 302
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5336828479796117,
      "learning_rate": 0.0009693362056920501,
      "loss": 3.166,
      "step": 303
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5456411100347327,
      "learning_rate": 0.0009690795997636015,
      "loss": 3.1777,
      "step": 304
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5093970967463356,
      "learning_rate": 0.0009688219588706179,
      "loss": 3.082,
      "step": 305
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.45893775260556363,
      "learning_rate": 0.0009685632835815518,
      "loss": 3.0488,
      "step": 306
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5466784760614037,
      "learning_rate": 0.0009683035744671367,
      "loss": 3.0176,
      "step": 307
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4749132203111788,
      "learning_rate": 0.0009680428321003883,
      "loss": 3.0977,
      "step": 308
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.43414839779237513,
      "learning_rate": 0.000967781057056601,
      "loss": 2.9922,
      "step": 309
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.569251649705193,
      "learning_rate": 0.0009675182499133485,
      "loss": 3.0527,
      "step": 310
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4379438532195839,
      "learning_rate": 0.0009672544112504813,
      "loss": 3.0664,
      "step": 311
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4809209003861618,
      "learning_rate": 0.0009669895416501257,
      "loss": 3.1855,
      "step": 312
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4033552190341695,
      "learning_rate": 0.0009667236416966833,
      "loss": 3.0898,
      "step": 313
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5219075609637426,
      "learning_rate": 0.0009664567119768281,
      "loss": 2.9121,
      "step": 314
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.40429439702648107,
      "learning_rate": 0.0009661887530795067,
      "loss": 3.0195,
      "step": 315
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.40665015115424774,
      "learning_rate": 0.0009659197655959365,
      "loss": 3.1094,
      "step": 316
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.390228833605134,
      "learning_rate": 0.000965649750119604,
      "loss": 3.0371,
      "step": 317
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4617044004561246,
      "learning_rate": 0.0009653787072462643,
      "loss": 3.1348,
      "step": 318
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4593345445791577,
      "learning_rate": 0.0009651066375739388,
      "loss": 3.0898,
      "step": 319
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.48927593539678726,
      "learning_rate": 0.000964833541702915,
      "loss": 3.1504,
      "step": 320
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.48054971436187277,
      "learning_rate": 0.0009645594202357438,
      "loss": 3.1172,
      "step": 321
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5726871869308142,
      "learning_rate": 0.0009642842737772397,
      "loss": 2.9629,
      "step": 322
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4869558468601563,
      "learning_rate": 0.0009640081029344782,
      "loss": 3.1016,
      "step": 323
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4884035429993351,
      "learning_rate": 0.0009637309083167956,
      "loss": 2.998,
      "step": 324
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5890894371282069,
      "learning_rate": 0.0009634526905357859,
      "loss": 3.0742,
      "step": 325
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4945797361375685,
      "learning_rate": 0.000963173450205302,
      "loss": 3.0996,
      "step": 326
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.44913688767669885,
      "learning_rate": 0.0009628931879414517,
      "loss": 3.1641,
      "step": 327
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5702501468414571,
      "learning_rate": 0.0009626119043625983,
      "loss": 3.1523,
      "step": 328
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4565559896611437,
      "learning_rate": 0.0009623296000893582,
      "loss": 3.0645,
      "step": 329
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4843216690527569,
      "learning_rate": 0.0009620462757446,
      "loss": 3.1445,
      "step": 330
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5024242074669293,
      "learning_rate": 0.0009617619319534428,
      "loss": 3.168,
      "step": 331
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.483474759497212,
      "learning_rate": 0.000961476569343255,
      "loss": 3.1328,
      "step": 332
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.40012488969136806,
      "learning_rate": 0.0009611901885436529,
      "loss": 3.0293,
      "step": 333
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.448000823257247,
      "learning_rate": 0.0009609027901864996,
      "loss": 3.0684,
      "step": 334
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4949904029741651,
      "learning_rate": 0.0009606143749059029,
      "loss": 3.1699,
      "step": 335
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5295588921000111,
      "learning_rate": 0.0009603249433382144,
      "loss": 3.0586,
      "step": 336
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5252475818003733,
      "learning_rate": 0.0009600344961220282,
      "loss": 3.0488,
      "step": 337
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4126149224020558,
      "learning_rate": 0.0009597430338981791,
      "loss": 3.0605,
      "step": 338
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.42274282780643374,
      "learning_rate": 0.0009594505573097414,
      "loss": 3.1426,
      "step": 339
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.47140613601958675,
      "learning_rate": 0.0009591570670020277,
      "loss": 3.0762,
      "step": 340
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4797127443685281,
      "learning_rate": 0.0009588625636225871,
      "loss": 3.0918,
      "step": 341
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4621357690988,
      "learning_rate": 0.0009585670478212036,
      "loss": 3.0977,
      "step": 342
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5448043453859415,
      "learning_rate": 0.0009582705202498956,
      "loss": 3.125,
      "step": 343
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4780856928317097,
      "learning_rate": 0.0009579729815629133,
      "loss": 3.1035,
      "step": 344
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4866978348536139,
      "learning_rate": 0.0009576744324167379,
      "loss": 2.9844,
      "step": 345
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.46798756418738985,
      "learning_rate": 0.0009573748734700804,
      "loss": 3.0605,
      "step": 346
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.47866728168162276,
      "learning_rate": 0.0009570743053838796,
      "loss": 3.0527,
      "step": 347
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4485656002772542,
      "learning_rate": 0.0009567727288213005,
      "loss": 3.168,
      "step": 348
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4502725029803653,
      "learning_rate": 0.0009564701444477337,
      "loss": 3.0469,
      "step": 349
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4115552324720369,
      "learning_rate": 0.000956166552930793,
      "loss": 3.0801,
      "step": 350
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.45577327035541665,
      "learning_rate": 0.0009558619549403147,
      "loss": 3.0957,
      "step": 351
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4594661642226814,
      "learning_rate": 0.0009555563511483555,
      "loss": 3.0859,
      "step": 352
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.40191081424283354,
      "learning_rate": 0.0009552497422291912,
      "loss": 3.2227,
      "step": 353
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.49678838563320277,
      "learning_rate": 0.0009549421288593157,
      "loss": 3.0352,
      "step": 354
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4454287338867608,
      "learning_rate": 0.0009546335117174385,
      "loss": 3.0957,
      "step": 355
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.44029323970763984,
      "learning_rate": 0.0009543238914844843,
      "loss": 3.043,
      "step": 356
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4220640644603041,
      "learning_rate": 0.0009540132688435907,
      "loss": 3.0859,
      "step": 357
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4038120157952765,
      "learning_rate": 0.0009537016444801074,
      "loss": 2.9961,
      "step": 358
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3855251799248939,
      "learning_rate": 0.0009533890190815935,
      "loss": 3.0332,
      "step": 359
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4092828730503744,
      "learning_rate": 0.0009530753933378173,
      "loss": 3.0703,
      "step": 360
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.48412823571149716,
      "learning_rate": 0.0009527607679407545,
      "loss": 3.0039,
      "step": 361
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4289039435362913,
      "learning_rate": 0.0009524451435845857,
      "loss": 3.0371,
      "step": 362
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5087026070260399,
      "learning_rate": 0.0009521285209656963,
      "loss": 3.1055,
      "step": 363
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.43150829924506023,
      "learning_rate": 0.0009518109007826734,
      "loss": 3.0,
      "step": 364
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.44957345877703,
      "learning_rate": 0.0009514922837363059,
      "loss": 3.0215,
      "step": 365
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4932974147910674,
      "learning_rate": 0.0009511726705295817,
      "loss": 3.0,
      "step": 366
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4057457975959084,
      "learning_rate": 0.000950852061867687,
      "loss": 3.0566,
      "step": 367
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4781755510272332,
      "learning_rate": 0.0009505304584580038,
      "loss": 3.1562,
      "step": 368
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4459979556401686,
      "learning_rate": 0.0009502078610101092,
      "loss": 3.1113,
      "step": 369
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.38422779112679195,
      "learning_rate": 0.0009498842702357736,
      "loss": 3.084,
      "step": 370
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.40814031622957514,
      "learning_rate": 0.0009495596868489587,
      "loss": 3.082,
      "step": 371
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.48211352319150585,
      "learning_rate": 0.0009492341115658167,
      "loss": 3.0801,
      "step": 372
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5453207426740636,
      "learning_rate": 0.0009489075451046879,
      "loss": 3.2031,
      "step": 373
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.44050154485037146,
      "learning_rate": 0.0009485799881861,
      "loss": 3.1543,
      "step": 374
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.461982796222206,
      "learning_rate": 0.0009482514415327654,
      "loss": 3.1777,
      "step": 375
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5255078470857125,
      "learning_rate": 0.000947921905869581,
      "loss": 3.0352,
      "step": 376
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.42894626251762863,
      "learning_rate": 0.0009475913819236248,
      "loss": 3.1543,
      "step": 377
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.44521179357481344,
      "learning_rate": 0.0009472598704241561,
      "loss": 3.0781,
      "step": 378
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.39396130428778064,
      "learning_rate": 0.0009469273721026131,
      "loss": 2.9746,
      "step": 379
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.48405620290402723,
      "learning_rate": 0.0009465938876926111,
      "loss": 3.0684,
      "step": 380
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.480794880250421,
      "learning_rate": 0.0009462594179299406,
      "loss": 3.0449,
      "step": 381
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4358404595645082,
      "learning_rate": 0.0009459239635525672,
      "loss": 3.0,
      "step": 382
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4992146336270293,
      "learning_rate": 0.0009455875253006281,
      "loss": 3.1191,
      "step": 383
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5136964409031926,
      "learning_rate": 0.0009452501039164315,
      "loss": 3.1367,
      "step": 384
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5252420874252131,
      "learning_rate": 0.0009449117001444549,
      "loss": 3.0879,
      "step": 385
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3965139910344841,
      "learning_rate": 0.0009445723147313433,
      "loss": 3.0254,
      "step": 386
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4777865005570892,
      "learning_rate": 0.0009442319484259074,
      "loss": 3.0527,
      "step": 387
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5211773664442485,
      "learning_rate": 0.0009438906019791222,
      "loss": 3.1367,
      "step": 388
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.43627002326766634,
      "learning_rate": 0.0009435482761441251,
      "loss": 3.1602,
      "step": 389
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5259868697286376,
      "learning_rate": 0.000943204971676215,
      "loss": 3.0254,
      "step": 390
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5774259023596658,
      "learning_rate": 0.0009428606893328493,
      "loss": 3.0801,
      "step": 391
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4634879991137712,
      "learning_rate": 0.0009425154298736432,
      "loss": 3.0312,
      "step": 392
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4517170525651034,
      "learning_rate": 0.0009421691940603678,
      "loss": 3.1094,
      "step": 393
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5345900651110781,
      "learning_rate": 0.0009418219826569488,
      "loss": 3.1582,
      "step": 394
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4109539894251694,
      "learning_rate": 0.0009414737964294635,
      "loss": 3.0684,
      "step": 395
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.44407594400113237,
      "learning_rate": 0.000941124636146141,
      "loss": 3.1113,
      "step": 396
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4002210444029189,
      "learning_rate": 0.0009407745025773589,
      "loss": 3.0605,
      "step": 397
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.46418238421475944,
      "learning_rate": 0.0009404233964956423,
      "loss": 3.0078,
      "step": 398
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.47983251482802675,
      "learning_rate": 0.0009400713186756625,
      "loss": 3.0117,
      "step": 399
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4587554821253575,
      "learning_rate": 0.0009397182698942342,
      "loss": 3.0938,
      "step": 400
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4474435726818662,
      "learning_rate": 0.0009393642509303149,
      "loss": 3.0156,
      "step": 401
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4738947589487815,
      "learning_rate": 0.0009390092625650023,
      "loss": 3.0977,
      "step": 402
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3848809292468056,
      "learning_rate": 0.0009386533055815332,
      "loss": 2.957,
      "step": 403
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5044233743131037,
      "learning_rate": 0.0009382963807652813,
      "loss": 3.1152,
      "step": 404
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5472034154744869,
      "learning_rate": 0.000937938488903756,
      "loss": 3.0469,
      "step": 405
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4217317019639569,
      "learning_rate": 0.0009375796307866003,
      "loss": 3.0957,
      "step": 406
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4721840810837384,
      "learning_rate": 0.0009372198072055888,
      "loss": 3.0762,
      "step": 407
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.48631442528386454,
      "learning_rate": 0.0009368590189546268,
      "loss": 3.0332,
      "step": 408
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4255291260735353,
      "learning_rate": 0.0009364972668297474,
      "loss": 3.0527,
      "step": 409
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5203766161511927,
      "learning_rate": 0.0009361345516291111,
      "loss": 3.0664,
      "step": 410
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.48895394743883097,
      "learning_rate": 0.0009357708741530024,
      "loss": 3.0371,
      "step": 411
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.40540951064187797,
      "learning_rate": 0.00093540623520383,
      "loss": 2.9512,
      "step": 412
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4692050341394939,
      "learning_rate": 0.000935040635586123,
      "loss": 3.0938,
      "step": 413
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4258128215021085,
      "learning_rate": 0.0009346740761065305,
      "loss": 3.1348,
      "step": 414
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4497921722922327,
      "learning_rate": 0.0009343065575738197,
      "loss": 3.002,
      "step": 415
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5469159606405699,
      "learning_rate": 0.0009339380807988733,
      "loss": 3.1406,
      "step": 416
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4793427436216943,
      "learning_rate": 0.0009335686465946887,
      "loss": 3.0176,
      "step": 417
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.43683129137627236,
      "learning_rate": 0.0009331982557763754,
      "loss": 2.9863,
      "step": 418
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.46512132953722285,
      "learning_rate": 0.0009328269091611537,
      "loss": 3.0039,
      "step": 419
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5037970768583495,
      "learning_rate": 0.0009324546075683524,
      "loss": 3.041,
      "step": 420
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4540677014731651,
      "learning_rate": 0.0009320813518194083,
      "loss": 3.0996,
      "step": 421
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5184653207347876,
      "learning_rate": 0.0009317071427378624,
      "loss": 3.0352,
      "step": 422
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5168489718818099,
      "learning_rate": 0.0009313319811493594,
      "loss": 3.0879,
      "step": 423
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.42383237757779707,
      "learning_rate": 0.000930955867881646,
      "loss": 3.0273,
      "step": 424
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5212508082713213,
      "learning_rate": 0.0009305788037645681,
      "loss": 3.0332,
      "step": 425
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.49093504835771096,
      "learning_rate": 0.0009302007896300697,
      "loss": 3.0293,
      "step": 426
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5255887013696354,
      "learning_rate": 0.0009298218263121911,
      "loss": 3.0254,
      "step": 427
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4182541949520963,
      "learning_rate": 0.0009294419146470668,
      "loss": 3.0664,
      "step": 428
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5051402705975846,
      "learning_rate": 0.0009290610554729234,
      "loss": 3.0801,
      "step": 429
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4358426296398898,
      "learning_rate": 0.0009286792496300784,
      "loss": 3.0488,
      "step": 430
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5465125492678916,
      "learning_rate": 0.0009282964979609379,
      "loss": 3.0625,
      "step": 431
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.47001020181933817,
      "learning_rate": 0.0009279128013099947,
      "loss": 3.0293,
      "step": 432
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.39576341942541093,
      "learning_rate": 0.0009275281605238268,
      "loss": 3.1504,
      "step": 433
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4760088090100082,
      "learning_rate": 0.0009271425764510953,
      "loss": 2.9883,
      "step": 434
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.45881998286716036,
      "learning_rate": 0.0009267560499425423,
      "loss": 3.0,
      "step": 435
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4586634461742784,
      "learning_rate": 0.0009263685818509895,
      "loss": 3.1543,
      "step": 436
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4416080376654591,
      "learning_rate": 0.000925980173031336,
      "loss": 3.0684,
      "step": 437
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.44388859191751306,
      "learning_rate": 0.0009255908243405567,
      "loss": 3.0469,
      "step": 438
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3901521554321602,
      "learning_rate": 0.0009252005366376996,
      "loss": 3.0156,
      "step": 439
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4043747084925032,
      "learning_rate": 0.0009248093107838852,
      "loss": 2.9746,
      "step": 440
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.43840064305699783,
      "learning_rate": 0.0009244171476423036,
      "loss": 3.0117,
      "step": 441
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4087733132148827,
      "learning_rate": 0.0009240240480782129,
      "loss": 3.0059,
      "step": 442
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.45110111318989093,
      "learning_rate": 0.0009236300129589376,
      "loss": 3.1445,
      "step": 443
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.40579909068335546,
      "learning_rate": 0.0009232350431538657,
      "loss": 3.0664,
      "step": 444
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4954087638950558,
      "learning_rate": 0.0009228391395344482,
      "loss": 2.9863,
      "step": 445
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3722609121683466,
      "learning_rate": 0.000922442302974196,
      "loss": 3.043,
      "step": 446
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.47168204635100386,
      "learning_rate": 0.0009220445343486785,
      "loss": 3.0742,
      "step": 447
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4479916654620829,
      "learning_rate": 0.0009216458345355217,
      "loss": 3.082,
      "step": 448
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4003624555263173,
      "learning_rate": 0.0009212462044144061,
      "loss": 3.0488,
      "step": 449
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4636673523838795,
      "learning_rate": 0.0009208456448670648,
      "loss": 3.0742,
      "step": 450
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4933427695496305,
      "learning_rate": 0.0009204441567772816,
      "loss": 3.0645,
      "step": 451
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4418833025801847,
      "learning_rate": 0.0009200417410308888,
      "loss": 3.0859,
      "step": 452
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4802838205908133,
      "learning_rate": 0.0009196383985157656,
      "loss": 2.9648,
      "step": 453
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5440162656235844,
      "learning_rate": 0.000919234130121836,
      "loss": 3.0762,
      "step": 454
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.45983292671636883,
      "learning_rate": 0.0009188289367410672,
      "loss": 3.0508,
      "step": 455
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5413532560528771,
      "learning_rate": 0.0009184228192674666,
      "loss": 3.0332,
      "step": 456
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4002004907574617,
      "learning_rate": 0.0009180157785970808,
      "loss": 2.998,
      "step": 457
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.539253110218124,
      "learning_rate": 0.0009176078156279932,
      "loss": 3.0996,
      "step": 458
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5454423401544124,
      "learning_rate": 0.0009171989312603226,
      "loss": 3.0371,
      "step": 459
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5449216967468329,
      "learning_rate": 0.0009167891263962202,
      "loss": 2.9883,
      "step": 460
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.44401285109168376,
      "learning_rate": 0.0009163784019398685,
      "loss": 2.9199,
      "step": 461
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4926757115143328,
      "learning_rate": 0.0009159667587974785,
      "loss": 3.0859,
      "step": 462
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5250590198139296,
      "learning_rate": 0.0009155541978772887,
      "loss": 3.0605,
      "step": 463
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5868334090474294,
      "learning_rate": 0.0009151407200895625,
      "loss": 3.0527,
      "step": 464
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.48116053590982577,
      "learning_rate": 0.000914726326346586,
      "loss": 3.0312,
      "step": 465
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.47207564837160115,
      "learning_rate": 0.0009143110175626661,
      "loss": 3.0156,
      "step": 466
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4536012029361483,
      "learning_rate": 0.0009138947946541291,
      "loss": 3.0,
      "step": 467
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.466508175295549,
      "learning_rate": 0.0009134776585393181,
      "loss": 3.082,
      "step": 468
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.49052790117948847,
      "learning_rate": 0.0009130596101385906,
      "loss": 3.002,
      "step": 469
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5378826433907979,
      "learning_rate": 0.0009126406503743174,
      "loss": 2.9707,
      "step": 470
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4754517450550239,
      "learning_rate": 0.0009122207801708802,
      "loss": 2.9805,
      "step": 471
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.48272088935953095,
      "learning_rate": 0.0009118000004546689,
      "loss": 3.0859,
      "step": 472
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4446622468924122,
      "learning_rate": 0.0009113783121540807,
      "loss": 3.0215,
      "step": 473
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4851776141798477,
      "learning_rate": 0.0009109557161995172,
      "loss": 3.0352,
      "step": 474
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5032561907230727,
      "learning_rate": 0.0009105322135233828,
      "loss": 3.0449,
      "step": 475
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.48388382677331315,
      "learning_rate": 0.0009101078050600821,
      "loss": 3.1035,
      "step": 476
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4420679906678583,
      "learning_rate": 0.0009096824917460186,
      "loss": 3.002,
      "step": 477
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.49675060448515046,
      "learning_rate": 0.0009092562745195921,
      "loss": 3.0215,
      "step": 478
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4167864762890616,
      "learning_rate": 0.0009088291543211967,
      "loss": 3.0488,
      "step": 479
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.46188405770663077,
      "learning_rate": 0.0009084011320932188,
      "loss": 3.0332,
      "step": 480
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.449153066062378,
      "learning_rate": 0.0009079722087800352,
      "loss": 3.0547,
      "step": 481
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4300827629316388,
      "learning_rate": 0.0009075423853280106,
      "loss": 3.0742,
      "step": 482
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4054383926386117,
      "learning_rate": 0.0009071116626854958,
      "loss": 3.0801,
      "step": 483
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4516065862499803,
      "learning_rate": 0.0009066800418028256,
      "loss": 3.0332,
      "step": 484
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.566202123702447,
      "learning_rate": 0.0009062475236323168,
      "loss": 3.082,
      "step": 485
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.44387901620139103,
      "learning_rate": 0.0009058141091282656,
      "loss": 2.9902,
      "step": 486
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.414085396160897,
      "learning_rate": 0.0009053797992469461,
      "loss": 3.0391,
      "step": 487
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.43646465349477037,
      "learning_rate": 0.0009049445949466078,
      "loss": 3.0488,
      "step": 488
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4705993387799091,
      "learning_rate": 0.0009045084971874737,
      "loss": 3.0801,
      "step": 489
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.41703024088656,
      "learning_rate": 0.0009040715069317382,
      "loss": 3.0586,
      "step": 490
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5161297578649258,
      "learning_rate": 0.0009036336251435648,
      "loss": 2.9785,
      "step": 491
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4853380289987665,
      "learning_rate": 0.0009031948527890839,
      "loss": 3.0488,
      "step": 492
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.486337893989459,
      "learning_rate": 0.000902755190836391,
      "loss": 3.0703,
      "step": 493
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.44525482684350587,
      "learning_rate": 0.0009023146402555442,
      "loss": 2.9766,
      "step": 494
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6269613366001471,
      "learning_rate": 0.0009018732020185624,
      "loss": 3.0547,
      "step": 495
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5308028372337257,
      "learning_rate": 0.0009014308770994235,
      "loss": 3.1016,
      "step": 496
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5399617654803657,
      "learning_rate": 0.0009009876664740605,
      "loss": 3.0156,
      "step": 497
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5470294716402231,
      "learning_rate": 0.0009005435711203618,
      "loss": 3.0762,
      "step": 498
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4250961100892837,
      "learning_rate": 0.000900098592018167,
      "loss": 2.9746,
      "step": 499
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4834582594620439,
      "learning_rate": 0.0008996527301492663,
      "loss": 3.082,
      "step": 500
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4324913052492657,
      "learning_rate": 0.0008992059864973972,
      "loss": 3.0352,
      "step": 501
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4123992683858629,
      "learning_rate": 0.0008987583620482427,
      "loss": 3.0898,
      "step": 502
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.41442166114915063,
      "learning_rate": 0.0008983098577894292,
      "loss": 2.9785,
      "step": 503
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.47902956955447773,
      "learning_rate": 0.0008978604747105246,
      "loss": 3.0684,
      "step": 504
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4638346972506078,
      "learning_rate": 0.0008974102138030354,
      "loss": 2.9395,
      "step": 505
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.39848892893352633,
      "learning_rate": 0.000896959076060405,
      "loss": 3.0723,
      "step": 506
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3757470362507673,
      "learning_rate": 0.0008965070624780116,
      "loss": 2.9824,
      "step": 507
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4057690636111222,
      "learning_rate": 0.0008960541740531658,
      "loss": 2.9238,
      "step": 508
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.42669342678393196,
      "learning_rate": 0.0008956004117851083,
      "loss": 3.0488,
      "step": 509
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4417887723474937,
      "learning_rate": 0.0008951457766750079,
      "loss": 3.0254,
      "step": 510
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4338396363801506,
      "learning_rate": 0.0008946902697259593,
      "loss": 3.0547,
      "step": 511
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.40525553732717556,
      "learning_rate": 0.0008942338919429805,
      "loss": 3.0488,
      "step": 512
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.47115511362240425,
      "learning_rate": 0.0008937766443330113,
      "loss": 3.0566,
      "step": 513
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.46415921549387706,
      "learning_rate": 0.0008933185279049103,
      "loss": 2.9492,
      "step": 514
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.46023773076255464,
      "learning_rate": 0.0008928595436694532,
      "loss": 3.041,
      "step": 515
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4110283114449815,
      "learning_rate": 0.0008923996926393305,
      "loss": 3.1094,
      "step": 516
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4454864740971105,
      "learning_rate": 0.0008919389758291449,
      "loss": 3.0801,
      "step": 517
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.41939399765690577,
      "learning_rate": 0.0008914773942554098,
      "loss": 3.0117,
      "step": 518
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.500922314543379,
      "learning_rate": 0.000891014948936546,
      "loss": 3.0977,
      "step": 519
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4621071469586475,
      "learning_rate": 0.0008905516408928804,
      "loss": 2.9727,
      "step": 520
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.42103344598651193,
      "learning_rate": 0.0008900874711466434,
      "loss": 3.0625,
      "step": 521
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.37333974569309664,
      "learning_rate": 0.0008896224407219666,
      "loss": 2.9434,
      "step": 522
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4031595922358347,
      "learning_rate": 0.0008891565506448804,
      "loss": 3.0371,
      "step": 523
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.48094071727280985,
      "learning_rate": 0.0008886898019433122,
      "loss": 2.9922,
      "step": 524
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.45356084082402437,
      "learning_rate": 0.0008882221956470836,
      "loss": 2.9551,
      "step": 525
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.449863471538791,
      "learning_rate": 0.0008877537327879086,
      "loss": 2.9707,
      "step": 526
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.536821700727635,
      "learning_rate": 0.0008872844143993908,
      "loss": 2.9941,
      "step": 527
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4227488654218183,
      "learning_rate": 0.0008868142415170218,
      "loss": 2.9551,
      "step": 528
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4392271676561706,
      "learning_rate": 0.0008863432151781781,
      "loss": 3.0215,
      "step": 529
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4871839597313624,
      "learning_rate": 0.0008858713364221195,
      "loss": 2.918,
      "step": 530
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5102270551055008,
      "learning_rate": 0.0008853986062899868,
      "loss": 3.0449,
      "step": 531
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.525537845517045,
      "learning_rate": 0.0008849250258247986,
      "loss": 3.0215,
      "step": 532
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5163602569064868,
      "learning_rate": 0.0008844505960714503,
      "loss": 3.0117,
      "step": 533
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.46997919889301387,
      "learning_rate": 0.0008839753180767108,
      "loss": 3.0332,
      "step": 534
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4385494408106652,
      "learning_rate": 0.0008834991928892204,
      "loss": 3.0215,
      "step": 535
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.6068071287690888,
      "learning_rate": 0.000883022221559489,
      "loss": 3.0098,
      "step": 536
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4769139700706598,
      "learning_rate": 0.0008825444051398934,
      "loss": 3.0879,
      "step": 537
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4275287553153966,
      "learning_rate": 0.0008820657446846745,
      "loss": 3.0762,
      "step": 538
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4947110199835254,
      "learning_rate": 0.000881586241249936,
      "loss": 3.043,
      "step": 539
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5109511872873971,
      "learning_rate": 0.0008811058958936411,
      "loss": 2.9922,
      "step": 540
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4860963908441498,
      "learning_rate": 0.000880624709675611,
      "loss": 3.0449,
      "step": 541
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4274126332493764,
      "learning_rate": 0.000880142683657522,
      "loss": 2.998,
      "step": 542
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4301976572559592,
      "learning_rate": 0.0008796598189029029,
      "loss": 3.0527,
      "step": 543
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4349869166384215,
      "learning_rate": 0.0008791761164771338,
      "loss": 3.0547,
      "step": 544
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.47236802818959517,
      "learning_rate": 0.0008786915774474424,
      "loss": 2.9844,
      "step": 545
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.44834218190260056,
      "learning_rate": 0.0008782062028829027,
      "loss": 3.0684,
      "step": 546
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4356011163262734,
      "learning_rate": 0.0008777199938544318,
      "loss": 3.0449,
      "step": 547
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5044068096573778,
      "learning_rate": 0.0008772329514347883,
      "loss": 2.9727,
      "step": 548
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.437770012248833,
      "learning_rate": 0.0008767450766985694,
      "loss": 3.0273,
      "step": 549
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.41783109724876966,
      "learning_rate": 0.0008762563707222086,
      "loss": 3.0312,
      "step": 550
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5114093448394831,
      "learning_rate": 0.0008757668345839738,
      "loss": 3.0938,
      "step": 551
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.53386395820561,
      "learning_rate": 0.0008752764693639638,
      "loss": 3.0977,
      "step": 552
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.40482604212663054,
      "learning_rate": 0.0008747852761441078,
      "loss": 2.9258,
      "step": 553
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4356156316494839,
      "learning_rate": 0.0008742932560081607,
      "loss": 3.0,
      "step": 554
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4103028143415426,
      "learning_rate": 0.0008738004100417025,
      "loss": 2.9668,
      "step": 555
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.40349517518699307,
      "learning_rate": 0.0008733067393321355,
      "loss": 3.0,
      "step": 556
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.47937159031699184,
      "learning_rate": 0.000872812244968681,
      "loss": 3.0918,
      "step": 557
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.43838030898577623,
      "learning_rate": 0.0008723169280423783,
      "loss": 3.0234,
      "step": 558
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5139230566267331,
      "learning_rate": 0.0008718207896460811,
      "loss": 3.0,
      "step": 559
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4001982932040262,
      "learning_rate": 0.0008713238308744557,
      "loss": 3.0508,
      "step": 560
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.42461661004775475,
      "learning_rate": 0.0008708260528239789,
      "loss": 2.9785,
      "step": 561
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5051277917089365,
      "learning_rate": 0.000870327456592934,
      "loss": 3.084,
      "step": 562
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4485391133515597,
      "learning_rate": 0.0008698280432814107,
      "loss": 3.0312,
      "step": 563
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.44484654850374916,
      "learning_rate": 0.000869327813991301,
      "loss": 3.0273,
      "step": 564
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.46770952307974634,
      "learning_rate": 0.0008688267698262971,
      "loss": 2.9824,
      "step": 565
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4769932362971195,
      "learning_rate": 0.0008683249118918894,
      "loss": 2.9961,
      "step": 566
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.39782241109284733,
      "learning_rate": 0.0008678222412953637,
      "loss": 2.9824,
      "step": 567
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.405676178969055,
      "learning_rate": 0.0008673187591457987,
      "loss": 3.0039,
      "step": 568
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5111752037462834,
      "learning_rate": 0.0008668144665540639,
      "loss": 3.041,
      "step": 569
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.39298184322632534,
      "learning_rate": 0.0008663093646328167,
      "loss": 2.9883,
      "step": 570
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.528011212126553,
      "learning_rate": 0.0008658034544965003,
      "loss": 2.9512,
      "step": 571
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5298243773462753,
      "learning_rate": 0.0008652967372613412,
      "loss": 3.1035,
      "step": 572
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4522812203697159,
      "learning_rate": 0.0008647892140453466,
      "loss": 3.0508,
      "step": 573
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.42496935955652165,
      "learning_rate": 0.0008642808859683021,
      "loss": 2.9512,
      "step": 574
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5326914539313761,
      "learning_rate": 0.0008637717541517689,
      "loss": 3.0371,
      "step": 575
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5539278986333595,
      "learning_rate": 0.0008632618197190816,
      "loss": 3.0469,
      "step": 576
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4930638015154211,
      "learning_rate": 0.0008627510837953458,
      "loss": 3.0527,
      "step": 577
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5320054927762757,
      "learning_rate": 0.0008622395475074355,
      "loss": 2.9316,
      "step": 578
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5694206334132664,
      "learning_rate": 0.0008617272119839903,
      "loss": 3.0312,
      "step": 579
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4224950694361083,
      "learning_rate": 0.0008612140783554136,
      "loss": 3.0508,
      "step": 580
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4681967237026844,
      "learning_rate": 0.0008607001477538696,
      "loss": 2.9121,
      "step": 581
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5431857482419087,
      "learning_rate": 0.0008601854213132807,
      "loss": 3.0215,
      "step": 582
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4814444004755299,
      "learning_rate": 0.0008596699001693256,
      "loss": 2.9473,
      "step": 583
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.42833185686175806,
      "learning_rate": 0.000859153585459436,
      "loss": 3.0527,
      "step": 584
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5608956816300757,
      "learning_rate": 0.0008586364783227949,
      "loss": 3.0234,
      "step": 585
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5136551960238012,
      "learning_rate": 0.0008581185799003332,
      "loss": 2.9805,
      "step": 586
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.383563869992909,
      "learning_rate": 0.0008575998913347283,
      "loss": 2.9609,
      "step": 587
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.47124938907655034,
      "learning_rate": 0.0008570804137704004,
      "loss": 2.9512,
      "step": 588
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4273091433188409,
      "learning_rate": 0.0008565601483535108,
      "loss": 2.9648,
      "step": 589
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4262932467185918,
      "learning_rate": 0.0008560390962319591,
      "loss": 2.998,
      "step": 590
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4509790627672309,
      "learning_rate": 0.0008555172585553804,
      "loss": 3.0,
      "step": 591
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4576366194519733,
      "learning_rate": 0.0008549946364751435,
      "loss": 2.9941,
      "step": 592
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5371979095089443,
      "learning_rate": 0.0008544712311443475,
      "loss": 3.0312,
      "step": 593
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4540830284740682,
      "learning_rate": 0.0008539470437178196,
      "loss": 2.9727,
      "step": 594
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4525489393127494,
      "learning_rate": 0.000853422075352113,
      "loss": 3.0879,
      "step": 595
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.429760659590888,
      "learning_rate": 0.0008528963272055035,
      "loss": 3.0293,
      "step": 596
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5239026724973186,
      "learning_rate": 0.0008523698004379877,
      "loss": 3.0703,
      "step": 597
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4332243674254332,
      "learning_rate": 0.00085184249621128,
      "loss": 3.0078,
      "step": 598
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.45345912408457717,
      "learning_rate": 0.0008513144156888101,
      "loss": 3.0586,
      "step": 599
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4459687224585205,
      "learning_rate": 0.0008507855600357207,
      "loss": 2.9277,
      "step": 600
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.47600047986237054,
      "learning_rate": 0.0008502559304188644,
      "loss": 2.9648,
      "step": 601
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.476492848287245,
      "learning_rate": 0.0008497255280068019,
      "loss": 2.998,
      "step": 602
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.45052870277168056,
      "learning_rate": 0.0008491943539697986,
      "loss": 3.041,
      "step": 603
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4460422928685141,
      "learning_rate": 0.0008486624094798226,
      "loss": 3.0176,
      "step": 604
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4291292797152913,
      "learning_rate": 0.0008481296957105417,
      "loss": 3.1055,
      "step": 605
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.5189010192697251,
      "learning_rate": 0.0008475962138373213,
      "loss": 3.0547,
      "step": 606
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.44964344170343074,
      "learning_rate": 0.0008470619650372211,
      "loss": 3.0176,
      "step": 607
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.44677731163514656,
      "learning_rate": 0.0008465269504889934,
      "loss": 2.9414,
      "step": 608
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.45052284441855783,
      "learning_rate": 0.0008459911713730799,
      "loss": 3.1582,
      "step": 609
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4663915377828737,
      "learning_rate": 0.0008454546288716089,
      "loss": 2.9902,
      "step": 610
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.5091229426573425,
      "learning_rate": 0.0008449173241683935,
      "loss": 2.9961,
      "step": 611
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.46650304166617085,
      "learning_rate": 0.0008443792584489281,
      "loss": 2.9512,
      "step": 612
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4403597845703397,
      "learning_rate": 0.0008438404329003863,
      "loss": 2.9609,
      "step": 613
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.45409753304790806,
      "learning_rate": 0.0008433008487116183,
      "loss": 2.9941,
      "step": 614
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.47524241262362016,
      "learning_rate": 0.0008427605070731481,
      "loss": 3.0977,
      "step": 615
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4480780802537743,
      "learning_rate": 0.0008422194091771708,
      "loss": 2.9238,
      "step": 616
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4273916492135552,
      "learning_rate": 0.0008416775562175503,
      "loss": 2.9551,
      "step": 617
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.43135707520185373,
      "learning_rate": 0.000841134949389816,
      "loss": 2.9531,
      "step": 618
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.48185473753992836,
      "learning_rate": 0.0008405915898911611,
      "loss": 3.0371,
      "step": 619
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.49577867615340976,
      "learning_rate": 0.0008400474789204396,
      "loss": 2.998,
      "step": 620
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.445741251363541,
      "learning_rate": 0.0008395026176781626,
      "loss": 2.9863,
      "step": 621
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4122450866255194,
      "learning_rate": 0.0008389570073664976,
      "loss": 3.002,
      "step": 622
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.45187229922179684,
      "learning_rate": 0.0008384106491892642,
      "loss": 2.9414,
      "step": 623
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.48429285762417384,
      "learning_rate": 0.0008378635443519327,
      "loss": 2.9688,
      "step": 624
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.43145537018358626,
      "learning_rate": 0.0008373156940616199,
      "loss": 2.9414,
      "step": 625
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4917955980942865,
      "learning_rate": 0.0008367670995270882,
      "loss": 2.9629,
      "step": 626
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.44304102750496493,
      "learning_rate": 0.0008362177619587416,
      "loss": 3.0508,
      "step": 627
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4923028850586271,
      "learning_rate": 0.0008356676825686238,
      "loss": 2.9492,
      "step": 628
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4804914268515357,
      "learning_rate": 0.0008351168625704147,
      "loss": 3.0625,
      "step": 629
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.48714391292495873,
      "learning_rate": 0.0008345653031794292,
      "loss": 3.041,
      "step": 630
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.46904352824529993,
      "learning_rate": 0.0008340130056126125,
      "loss": 3.0625,
      "step": 631
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4818992261592087,
      "learning_rate": 0.0008334599710885394,
      "loss": 3.0938,
      "step": 632
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.5205958276155845,
      "learning_rate": 0.0008329062008274098,
      "loss": 3.0586,
      "step": 633
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.48460497177824474,
      "learning_rate": 0.000832351696051048,
      "loss": 2.9746,
      "step": 634
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4473704620011215,
      "learning_rate": 0.000831796457982898,
      "loss": 2.9766,
      "step": 635
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4801744673118726,
      "learning_rate": 0.0008312404878480222,
      "loss": 3.0156,
      "step": 636
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.486336866666704,
      "learning_rate": 0.0008306837868730979,
      "loss": 2.9375,
      "step": 637
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.438115967217394,
      "learning_rate": 0.0008301263562864152,
      "loss": 3.041,
      "step": 638
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.5402755495704495,
      "learning_rate": 0.0008295681973178737,
      "loss": 2.9297,
      "step": 639
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.5026726071450975,
      "learning_rate": 0.0008290093111989804,
      "loss": 3.0352,
      "step": 640
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4221588632350391,
      "learning_rate": 0.0008284496991628465,
      "loss": 2.8848,
      "step": 641
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4329529156466055,
      "learning_rate": 0.0008278893624441847,
      "loss": 3.0098,
      "step": 642
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.46010309235237007,
      "learning_rate": 0.000827328302279307,
      "loss": 2.998,
      "step": 643
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4536067346045542,
      "learning_rate": 0.0008267665199061211,
      "loss": 2.9668,
      "step": 644
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.45860683428055904,
      "learning_rate": 0.0008262040165641288,
      "loss": 3.0664,
      "step": 645
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4456081254489809,
      "learning_rate": 0.0008256407934944219,
      "loss": 2.9316,
      "step": 646
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4739683556637825,
      "learning_rate": 0.0008250768519396807,
      "loss": 2.9531,
      "step": 647
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4295056240266932,
      "learning_rate": 0.0008245121931441706,
      "loss": 2.9766,
      "step": 648
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.43915663734441357,
      "learning_rate": 0.0008239468183537393,
      "loss": 2.9219,
      "step": 649
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.525007141735391,
      "learning_rate": 0.0008233807288158146,
      "loss": 3.0332,
      "step": 650
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.41905293750895806,
      "learning_rate": 0.0008228139257794012,
      "loss": 2.9512,
      "step": 651
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4664712337649914,
      "learning_rate": 0.0008222464104950778,
      "loss": 3.082,
      "step": 652
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.40231705849468963,
      "learning_rate": 0.000821678184214995,
      "loss": 2.9746,
      "step": 653
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4208496349236011,
      "learning_rate": 0.0008211092481928716,
      "loss": 3.0156,
      "step": 654
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4234194813332057,
      "learning_rate": 0.0008205396036839927,
      "loss": 3.0215,
      "step": 655
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.44287275153804545,
      "learning_rate": 0.0008199692519452069,
      "loss": 3.0,
      "step": 656
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.375057642244709,
      "learning_rate": 0.0008193981942349224,
      "loss": 3.0117,
      "step": 657
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4554398759354132,
      "learning_rate": 0.0008188264318131056,
      "loss": 2.9727,
      "step": 658
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4579269230613815,
      "learning_rate": 0.0008182539659412776,
      "loss": 2.9824,
      "step": 659
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4533916300198751,
      "learning_rate": 0.0008176807978825118,
      "loss": 3.0098,
      "step": 660
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4990067943750579,
      "learning_rate": 0.0008171069289014306,
      "loss": 3.0195,
      "step": 661
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.49793881904882104,
      "learning_rate": 0.0008165323602642028,
      "loss": 2.9453,
      "step": 662
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.498257513067113,
      "learning_rate": 0.0008159570932385414,
      "loss": 3.0156,
      "step": 663
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5090561820979576,
      "learning_rate": 0.0008153811290936999,
      "loss": 3.0078,
      "step": 664
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3794810671909957,
      "learning_rate": 0.0008148044691004698,
      "loss": 2.9902,
      "step": 665
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.5006769398092179,
      "learning_rate": 0.0008142271145311783,
      "loss": 2.9746,
      "step": 666
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.46785954618480907,
      "learning_rate": 0.000813649066659685,
      "loss": 3.0293,
      "step": 667
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.44941828257451905,
      "learning_rate": 0.0008130703267613787,
      "loss": 3.0449,
      "step": 668
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.5223207331032744,
      "learning_rate": 0.0008124908961131759,
      "loss": 2.9785,
      "step": 669
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.45306953551115436,
      "learning_rate": 0.0008119107759935163,
      "loss": 3.0117,
      "step": 670
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4449499414958624,
      "learning_rate": 0.0008113299676823615,
      "loss": 2.9824,
      "step": 671
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.47973564772157246,
      "learning_rate": 0.0008107484724611911,
      "loss": 2.9629,
      "step": 672
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.47652651343625996,
      "learning_rate": 0.0008101662916130006,
      "loss": 2.9551,
      "step": 673
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4263433919041022,
      "learning_rate": 0.0008095834264222979,
      "loss": 2.9102,
      "step": 674
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4786161819150518,
      "learning_rate": 0.0008089998781751009,
      "loss": 3.0117,
      "step": 675
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.5612408061475225,
      "learning_rate": 0.0008084156481589349,
      "loss": 3.0625,
      "step": 676
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.46827549638780863,
      "learning_rate": 0.0008078307376628291,
      "loss": 3.0273,
      "step": 677
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.47889506190854425,
      "learning_rate": 0.0008072451479773143,
      "loss": 2.9668,
      "step": 678
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.43268816882424604,
      "learning_rate": 0.0008066588803944195,
      "loss": 3.0254,
      "step": 679
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4265177019078443,
      "learning_rate": 0.0008060719362076697,
      "loss": 2.9199,
      "step": 680
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4209561617810796,
      "learning_rate": 0.0008054843167120826,
      "loss": 2.9453,
      "step": 681
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4377556492895727,
      "learning_rate": 0.0008048960232041663,
      "loss": 3.0078,
      "step": 682
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.49312486560656943,
      "learning_rate": 0.0008043070569819153,
      "loss": 2.9473,
      "step": 683
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4529003066517036,
      "learning_rate": 0.0008037174193448089,
      "loss": 2.8926,
      "step": 684
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.5047606659915687,
      "learning_rate": 0.0008031271115938077,
      "loss": 3.0098,
      "step": 685
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.48439790296388563,
      "learning_rate": 0.0008025361350313505,
      "loss": 2.9863,
      "step": 686
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.5405864577718492,
      "learning_rate": 0.0008019444909613523,
      "loss": 3.0273,
      "step": 687
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4700890083466812,
      "learning_rate": 0.0008013521806892003,
      "loss": 2.9961,
      "step": 688
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.472969676425895,
      "learning_rate": 0.000800759205521752,
      "loss": 2.9199,
      "step": 689
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4415593375658549,
      "learning_rate": 0.0008001655667673318,
      "loss": 2.9336,
      "step": 690
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5552946142700617,
      "learning_rate": 0.0007995712657357279,
      "loss": 3.0684,
      "step": 691
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3918637208420732,
      "learning_rate": 0.0007989763037381904,
      "loss": 2.9824,
      "step": 692
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4663067999276016,
      "learning_rate": 0.0007983806820874271,
      "loss": 2.9805,
      "step": 693
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5047208497521289,
      "learning_rate": 0.0007977844020976016,
      "loss": 3.0977,
      "step": 694
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4748234593764113,
      "learning_rate": 0.00079718746508433,
      "loss": 2.9668,
      "step": 695
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4631403086038505,
      "learning_rate": 0.0007965898723646776,
      "loss": 2.9824,
      "step": 696
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4193796221227481,
      "learning_rate": 0.0007959916252571573,
      "loss": 2.8398,
      "step": 697
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5271327537056423,
      "learning_rate": 0.000795392725081725,
      "loss": 2.9785,
      "step": 698
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4445857642604608,
      "learning_rate": 0.000794793173159778,
      "loss": 2.9336,
      "step": 699
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4227914044354707,
      "learning_rate": 0.0007941929708141513,
      "loss": 2.9531,
      "step": 700
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4164366382689866,
      "learning_rate": 0.0007935921193691153,
      "loss": 2.9453,
      "step": 701
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4097851878690966,
      "learning_rate": 0.0007929906201503722,
      "loss": 2.9258,
      "step": 702
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5146963324668182,
      "learning_rate": 0.0007923884744850536,
      "loss": 3.0215,
      "step": 703
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5073753099787541,
      "learning_rate": 0.0007917856837017176,
      "loss": 2.9668,
      "step": 704
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.43413163047647396,
      "learning_rate": 0.0007911822491303452,
      "loss": 2.9863,
      "step": 705
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.42950971912337743,
      "learning_rate": 0.0007905781721023382,
      "loss": 2.9785,
      "step": 706
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4712457325048601,
      "learning_rate": 0.000789973453950516,
      "loss": 3.0352,
      "step": 707
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4505379992875525,
      "learning_rate": 0.000789368096009112,
      "loss": 2.9648,
      "step": 708
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.41548327626322373,
      "learning_rate": 0.0007887620996137721,
      "loss": 2.9141,
      "step": 709
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.45882209048286526,
      "learning_rate": 0.0007881554661015497,
      "loss": 3.0098,
      "step": 710
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4304727130037216,
      "learning_rate": 0.0007875481968109051,
      "loss": 2.9922,
      "step": 711
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4747298173626378,
      "learning_rate": 0.0007869402930817007,
      "loss": 2.9473,
      "step": 712
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4193645365524751,
      "learning_rate": 0.0007863317562551987,
      "loss": 2.9551,
      "step": 713
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4253240686801171,
      "learning_rate": 0.0007857225876740584,
      "loss": 3.0586,
      "step": 714
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4203541368548328,
      "learning_rate": 0.0007851127886823327,
      "loss": 2.8887,
      "step": 715
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.48035135521364264,
      "learning_rate": 0.0007845023606254658,
      "loss": 2.9434,
      "step": 716
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.5142778074243077,
      "learning_rate": 0.0007838913048502894,
      "loss": 2.9375,
      "step": 717
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.49298128795978696,
      "learning_rate": 0.0007832796227050208,
      "loss": 3.0078,
      "step": 718
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.46809222644529447,
      "learning_rate": 0.0007826673155392587,
      "loss": 3.0059,
      "step": 719
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.484751966221913,
      "learning_rate": 0.000782054384703981,
      "loss": 2.9727,
      "step": 720
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.46916544365184304,
      "learning_rate": 0.0007814408315515418,
      "loss": 2.9844,
      "step": 721
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4761046620868874,
      "learning_rate": 0.0007808266574356683,
      "loss": 2.9922,
      "step": 722
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.43426609236906283,
      "learning_rate": 0.0007802118637114573,
      "loss": 3.0,
      "step": 723
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4111717788534951,
      "learning_rate": 0.0007795964517353734,
      "loss": 2.957,
      "step": 724
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4485754730507227,
      "learning_rate": 0.0007789804228652449,
      "loss": 3.0176,
      "step": 725
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4097909957541059,
      "learning_rate": 0.0007783637784602609,
      "loss": 2.9707,
      "step": 726
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.46487666327721217,
      "learning_rate": 0.0007777465198809692,
      "loss": 2.9297,
      "step": 727
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.5113806880147794,
      "learning_rate": 0.0007771286484892722,
      "loss": 2.9824,
      "step": 728
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.45858672238337594,
      "learning_rate": 0.000776510165648425,
      "loss": 2.9375,
      "step": 729
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.4854411955915566,
      "learning_rate": 0.0007758910727230311,
      "loss": 2.9961,
      "step": 730
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.49380568079982656,
      "learning_rate": 0.0007752713710790404,
      "loss": 2.8633,
      "step": 731
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.43585170425451475,
      "learning_rate": 0.0007746510620837459,
      "loss": 2.9004,
      "step": 732
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4364677448076431,
      "learning_rate": 0.0007740301471057807,
      "loss": 2.9277,
      "step": 733
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4704403188736201,
      "learning_rate": 0.0007734086275151146,
      "loss": 2.9277,
      "step": 734
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4100138610093793,
      "learning_rate": 0.0007727865046830517,
      "loss": 2.9316,
      "step": 735
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.45999355498534167,
      "learning_rate": 0.0007721637799822269,
      "loss": 2.9434,
      "step": 736
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.46046857174390193,
      "learning_rate": 0.0007715404547866032,
      "loss": 2.9844,
      "step": 737
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.47578032252659974,
      "learning_rate": 0.0007709165304714685,
      "loss": 3.0117,
      "step": 738
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.457492456995111,
      "learning_rate": 0.0007702920084134324,
      "loss": 3.0039,
      "step": 739
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4859570135807596,
      "learning_rate": 0.0007696668899904236,
      "loss": 2.9297,
      "step": 740
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4288558968443924,
      "learning_rate": 0.0007690411765816864,
      "loss": 2.9375,
      "step": 741
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4203362237905022,
      "learning_rate": 0.0007684148695677778,
      "loss": 2.9258,
      "step": 742
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4085747161695624,
      "learning_rate": 0.000767787970330565,
      "loss": 2.9375,
      "step": 743
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.41227914850744857,
      "learning_rate": 0.000767160480253221,
      "loss": 2.9785,
      "step": 744
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4336664168068354,
      "learning_rate": 0.0007665324007202235,
      "loss": 2.9414,
      "step": 745
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.47194083170099754,
      "learning_rate": 0.0007659037331173498,
      "loss": 2.9004,
      "step": 746
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4537754648217551,
      "learning_rate": 0.0007652744788316752,
      "loss": 3.0449,
      "step": 747
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.427811483940314,
      "learning_rate": 0.0007646446392515692,
      "loss": 2.9121,
      "step": 748
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.5191171770313783,
      "learning_rate": 0.000764014215766693,
      "loss": 3.125,
      "step": 749
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.42960695069173016,
      "learning_rate": 0.0007633832097679958,
      "loss": 3.0078,
      "step": 750
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4533812064730918,
      "learning_rate": 0.0007627516226477122,
      "loss": 2.9219,
      "step": 751
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4589306006609965,
      "learning_rate": 0.0007621194557993589,
      "loss": 2.9609,
      "step": 752
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.45338298438679503,
      "learning_rate": 0.0007614867106177319,
      "loss": 2.9668,
      "step": 753
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4697459179661028,
      "learning_rate": 0.0007608533884989029,
      "loss": 3.0645,
      "step": 754
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4387826224753428,
      "learning_rate": 0.0007602194908402166,
      "loss": 3.0039,
      "step": 755
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.44530032689204135,
      "learning_rate": 0.0007595850190402877,
      "loss": 2.9941,
      "step": 756
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4402119893525058,
      "learning_rate": 0.0007589499744989976,
      "loss": 2.9395,
      "step": 757
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4629568450162128,
      "learning_rate": 0.0007583143586174916,
      "loss": 2.9512,
      "step": 758
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4816433893109519,
      "learning_rate": 0.000757678172798175,
      "loss": 2.9609,
      "step": 759
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4626287350609111,
      "learning_rate": 0.0007570414184447112,
      "loss": 2.9219,
      "step": 760
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4588318408982525,
      "learning_rate": 0.0007564040969620179,
      "loss": 2.9609,
      "step": 761
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4961179605108805,
      "learning_rate": 0.0007557662097562636,
      "loss": 2.9746,
      "step": 762
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4278181995624729,
      "learning_rate": 0.0007551277582348658,
      "loss": 2.9355,
      "step": 763
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4353033144418015,
      "learning_rate": 0.0007544887438064862,
      "loss": 3.002,
      "step": 764
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.5030726812256009,
      "learning_rate": 0.0007538491678810294,
      "loss": 3.0059,
      "step": 765
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4513450171381394,
      "learning_rate": 0.0007532090318696381,
      "loss": 2.9785,
      "step": 766
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4578011265372715,
      "learning_rate": 0.0007525683371846913,
      "loss": 2.959,
      "step": 767
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4862147938201292,
      "learning_rate": 0.0007519270852398001,
      "loss": 3.0039,
      "step": 768
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4707912294537616,
      "learning_rate": 0.000751285277449806,
      "loss": 3.0918,
      "step": 769
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4892872491409543,
      "learning_rate": 0.0007506429152307756,
      "loss": 3.082,
      "step": 770
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4286760369489084,
      "learning_rate": 0.00075,
      "loss": 2.9258,
      "step": 771
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.40467490305554843,
      "learning_rate": 0.00074935653317599,
      "loss": 2.8867,
      "step": 772
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.43951119694865065,
      "learning_rate": 0.000748712516178473,
      "loss": 2.9277,
      "step": 773
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.4546255026067793,
      "learning_rate": 0.0007480679504283911,
      "loss": 2.9258,
      "step": 774
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.45603313381710825,
      "learning_rate": 0.0007474228373478964,
      "loss": 2.9883,
      "step": 775
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.43531920998379936,
      "learning_rate": 0.0007467771783603492,
      "loss": 3.0137,
      "step": 776
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4662822782121286,
      "learning_rate": 0.0007461309748903138,
      "loss": 3.0098,
      "step": 777
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4469186265009868,
      "learning_rate": 0.0007454842283635562,
      "loss": 3.1152,
      "step": 778
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.5085960305043041,
      "learning_rate": 0.0007448369402070404,
      "loss": 2.8887,
      "step": 779
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.5096183368017347,
      "learning_rate": 0.0007441891118489254,
      "loss": 2.998,
      "step": 780
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4466138176316578,
      "learning_rate": 0.0007435407447185622,
      "loss": 3.0371,
      "step": 781
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4944650377932557,
      "learning_rate": 0.0007428918402464908,
      "loss": 2.9824,
      "step": 782
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.42846207650075946,
      "learning_rate": 0.0007422423998644359,
      "loss": 2.9434,
      "step": 783
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.5048417453335806,
      "learning_rate": 0.0007415924250053055,
      "loss": 2.9434,
      "step": 784
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4553482411954884,
      "learning_rate": 0.0007409419171031865,
      "loss": 2.9531,
      "step": 785
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4496341875684212,
      "learning_rate": 0.0007402908775933419,
      "loss": 2.9121,
      "step": 786
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4105199696223913,
      "learning_rate": 0.0007396393079122077,
      "loss": 2.9355,
      "step": 787
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.46162940104203914,
      "learning_rate": 0.0007389872094973896,
      "loss": 2.9766,
      "step": 788
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.49158517982765215,
      "learning_rate": 0.00073833458378766,
      "loss": 3.1152,
      "step": 789
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4222002872208981,
      "learning_rate": 0.0007376814322229544,
      "loss": 3.0684,
      "step": 790
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.48347276070499984,
      "learning_rate": 0.0007370277562443688,
      "loss": 3.0098,
      "step": 791
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.5094534887597232,
      "learning_rate": 0.0007363735572941564,
      "loss": 2.9375,
      "step": 792
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.47109644737798934,
      "learning_rate": 0.0007357188368157236,
      "loss": 3.0566,
      "step": 793
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4055813938530643,
      "learning_rate": 0.0007350635962536284,
      "loss": 2.8145,
      "step": 794
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.46626366026198857,
      "learning_rate": 0.0007344078370535756,
      "loss": 2.9688,
      "step": 795
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.4955015669041147,
      "learning_rate": 0.0007337515606624148,
      "loss": 3.0508,
      "step": 796
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.44550444532394623,
      "learning_rate": 0.0007330947685281362,
      "loss": 3.0332,
      "step": 797
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.43418147101451027,
      "learning_rate": 0.0007324374620998682,
      "loss": 3.0117,
      "step": 798
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.48018195755460896,
      "learning_rate": 0.000731779642827874,
      "loss": 2.8594,
      "step": 799
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4377517163330675,
      "learning_rate": 0.0007311213121635483,
      "loss": 2.959,
      "step": 800
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4662820724864323,
      "learning_rate": 0.0007304624715594139,
      "loss": 2.918,
      "step": 801
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.44896391188301027,
      "learning_rate": 0.0007298031224691193,
      "loss": 2.9766,
      "step": 802
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.45533597356012445,
      "learning_rate": 0.0007291432663474339,
      "loss": 2.9766,
      "step": 803
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4457541980474289,
      "learning_rate": 0.0007284829046502467,
      "loss": 2.9316,
      "step": 804
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4314556321788029,
      "learning_rate": 0.0007278220388345619,
      "loss": 2.8887,
      "step": 805
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.436166477812571,
      "learning_rate": 0.0007271606703584958,
      "loss": 2.9512,
      "step": 806
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.43534308083820744,
      "learning_rate": 0.000726498800681274,
      "loss": 2.9434,
      "step": 807
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4617840663561001,
      "learning_rate": 0.0007258364312632279,
      "loss": 2.9492,
      "step": 808
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4809007974247689,
      "learning_rate": 0.0007251735635657915,
      "loss": 2.9316,
      "step": 809
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5120727868957553,
      "learning_rate": 0.000724510199051498,
      "loss": 2.9004,
      "step": 810
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5070511796381881,
      "learning_rate": 0.0007238463391839769,
      "loss": 2.9961,
      "step": 811
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5104782416432423,
      "learning_rate": 0.0007231819854279508,
      "loss": 3.0156,
      "step": 812
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.486280041393241,
      "learning_rate": 0.0007225171392492316,
      "loss": 2.9102,
      "step": 813
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5533836176144116,
      "learning_rate": 0.0007218518021147182,
      "loss": 2.9922,
      "step": 814
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.5091366125403517,
      "learning_rate": 0.0007211859754923923,
      "loss": 2.9805,
      "step": 815
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.43998734517340043,
      "learning_rate": 0.0007205196608513158,
      "loss": 2.9863,
      "step": 816
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4021368722226169,
      "learning_rate": 0.0007198528596616272,
      "loss": 2.8418,
      "step": 817
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4713775463517793,
      "learning_rate": 0.0007191855733945387,
      "loss": 2.9297,
      "step": 818
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.5304257811970503,
      "learning_rate": 0.0007185178035223327,
      "loss": 2.9648,
      "step": 819
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.44787408776221593,
      "learning_rate": 0.0007178495515183583,
      "loss": 2.9863,
      "step": 820
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4787952366125969,
      "learning_rate": 0.000717180818857029,
      "loss": 2.9863,
      "step": 821
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.47150062151909916,
      "learning_rate": 0.0007165116070138182,
      "loss": 3.0625,
      "step": 822
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.451380652019378,
      "learning_rate": 0.0007158419174652569,
      "loss": 2.9453,
      "step": 823
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4657567914539754,
      "learning_rate": 0.00071517175168893,
      "loss": 2.9258,
      "step": 824
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4609226449761406,
      "learning_rate": 0.0007145011111634732,
      "loss": 2.9492,
      "step": 825
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4524997210465232,
      "learning_rate": 0.0007138299973685694,
      "loss": 3.0488,
      "step": 826
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4511387610501086,
      "learning_rate": 0.0007131584117849459,
      "loss": 2.9844,
      "step": 827
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.42792301911651953,
      "learning_rate": 0.0007124863558943713,
      "loss": 2.9043,
      "step": 828
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4492215697767553,
      "learning_rate": 0.0007118138311796514,
      "loss": 2.9453,
      "step": 829
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4700665331958391,
      "learning_rate": 0.0007111408391246262,
      "loss": 3.0332,
      "step": 830
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.5069396818588535,
      "learning_rate": 0.0007104673812141675,
      "loss": 2.9473,
      "step": 831
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.47325445288227036,
      "learning_rate": 0.0007097934589341745,
      "loss": 2.9453,
      "step": 832
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4866751524017525,
      "learning_rate": 0.0007091190737715711,
      "loss": 2.8926,
      "step": 833
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.42019172999972254,
      "learning_rate": 0.0007084442272143026,
      "loss": 2.9844,
      "step": 834
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.5107644015711499,
      "learning_rate": 0.000707768920751332,
      "loss": 3.0293,
      "step": 835
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4047599043392518,
      "learning_rate": 0.0007070931558726373,
      "loss": 2.9746,
      "step": 836
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4630564024843144,
      "learning_rate": 0.0007064169340692076,
      "loss": 2.9941,
      "step": 837
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4100151601299989,
      "learning_rate": 0.0007057402568330407,
      "loss": 3.0293,
      "step": 838
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.45227439157334515,
      "learning_rate": 0.0007050631256571389,
      "loss": 3.0293,
      "step": 839
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.459416026764158,
      "learning_rate": 0.000704385542035506,
      "loss": 2.959,
      "step": 840
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.46612242135970405,
      "learning_rate": 0.000703707507463144,
      "loss": 3.0254,
      "step": 841
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4383566326380167,
      "learning_rate": 0.0007030290234360505,
      "loss": 2.9453,
      "step": 842
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.44662335014056875,
      "learning_rate": 0.0007023500914512139,
      "loss": 2.9453,
      "step": 843
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4635480527992373,
      "learning_rate": 0.0007016707130066116,
      "loss": 2.9668,
      "step": 844
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.42385153431809286,
      "learning_rate": 0.0007009908896012055,
      "loss": 3.0,
      "step": 845
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4699004068996787,
      "learning_rate": 0.0007003106227349399,
      "loss": 2.9492,
      "step": 846
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.44642065406807735,
      "learning_rate": 0.000699629913908737,
      "loss": 2.9219,
      "step": 847
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4494267153984445,
      "learning_rate": 0.0006989487646244943,
      "loss": 2.8691,
      "step": 848
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4725026848739869,
      "learning_rate": 0.0006982671763850814,
      "loss": 2.9297,
      "step": 849
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.46248591251586557,
      "learning_rate": 0.0006975851506943359,
      "loss": 2.916,
      "step": 850
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.47728726004038796,
      "learning_rate": 0.0006969026890570611,
      "loss": 3.002,
      "step": 851
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.49536879260313016,
      "learning_rate": 0.0006962197929790216,
      "loss": 2.957,
      "step": 852
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4195300292240617,
      "learning_rate": 0.0006955364639669409,
      "loss": 2.9707,
      "step": 853
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.42474586683168275,
      "learning_rate": 0.0006948527035284978,
      "loss": 2.957,
      "step": 854
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4529525630667581,
      "learning_rate": 0.0006941685131723225,
      "loss": 2.9512,
      "step": 855
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.47823001852570085,
      "learning_rate": 0.0006934838944079943,
      "loss": 2.9785,
      "step": 856
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4117545570332013,
      "learning_rate": 0.0006927988487460378,
      "loss": 2.8848,
      "step": 857
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4648200080873134,
      "learning_rate": 0.0006921133776979186,
      "loss": 2.9531,
      "step": 858
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4153712439744175,
      "learning_rate": 0.0006914274827760418,
      "loss": 2.9355,
      "step": 859
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4435720205870345,
      "learning_rate": 0.0006907411654937475,
      "loss": 3.041,
      "step": 860
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.4370594987842312,
      "learning_rate": 0.0006900544273653075,
      "loss": 2.9453,
      "step": 861
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.40582041593444157,
      "learning_rate": 0.000689367269905922,
      "loss": 2.9805,
      "step": 862
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.47514586391541747,
      "learning_rate": 0.0006886796946317168,
      "loss": 2.9512,
      "step": 863
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4423354883215249,
      "learning_rate": 0.0006879917030597397,
      "loss": 2.9727,
      "step": 864
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.5261400288078218,
      "learning_rate": 0.0006873032967079561,
      "loss": 3.0137,
      "step": 865
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4956794003311002,
      "learning_rate": 0.0006866144770952474,
      "loss": 2.9395,
      "step": 866
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.43256075355899853,
      "learning_rate": 0.0006859252457414067,
      "loss": 2.9824,
      "step": 867
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.45125822902818546,
      "learning_rate": 0.0006852356041671351,
      "loss": 2.9844,
      "step": 868
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4543157409863931,
      "learning_rate": 0.0006845455538940394,
      "loss": 2.9941,
      "step": 869
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4235840046096599,
      "learning_rate": 0.0006838550964446276,
      "loss": 2.8906,
      "step": 870
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.44509020186514114,
      "learning_rate": 0.0006831642333423067,
      "loss": 2.9551,
      "step": 871
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4318333059266962,
      "learning_rate": 0.000682472966111378,
      "loss": 2.9199,
      "step": 872
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4392445045570088,
      "learning_rate": 0.0006817812962770348,
      "loss": 3.0078,
      "step": 873
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4982007531387423,
      "learning_rate": 0.0006810892253653589,
      "loss": 2.916,
      "step": 874
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4784443519084296,
      "learning_rate": 0.0006803967549033167,
      "loss": 2.9727,
      "step": 875
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.44675437403739926,
      "learning_rate": 0.0006797038864187564,
      "loss": 2.9805,
      "step": 876
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4618509759689776,
      "learning_rate": 0.0006790106214404043,
      "loss": 3.0059,
      "step": 877
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.46446288633355576,
      "learning_rate": 0.0006783169614978614,
      "loss": 2.9766,
      "step": 878
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4521755935197461,
      "learning_rate": 0.0006776229081216001,
      "loss": 2.9336,
      "step": 879
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.44485150807291357,
      "learning_rate": 0.0006769284628429611,
      "loss": 3.0098,
      "step": 880
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.44894151347834954,
      "learning_rate": 0.0006762336271941498,
      "loss": 3.0215,
      "step": 881
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.42108638302124884,
      "learning_rate": 0.0006755384027082326,
      "loss": 2.8691,
      "step": 882
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.46271429772715345,
      "learning_rate": 0.0006748427909191342,
      "loss": 2.9414,
      "step": 883
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.49728513078408404,
      "learning_rate": 0.0006741467933616335,
      "loss": 2.998,
      "step": 884
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4838933479921247,
      "learning_rate": 0.0006734504115713604,
      "loss": 2.8457,
      "step": 885
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4681600581771918,
      "learning_rate": 0.0006727536470847932,
      "loss": 2.9043,
      "step": 886
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4449965051288844,
      "learning_rate": 0.000672056501439254,
      "loss": 2.959,
      "step": 887
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.43272794925968,
      "learning_rate": 0.0006713589761729063,
      "loss": 2.918,
      "step": 888
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4415240425338209,
      "learning_rate": 0.0006706610728247508,
      "loss": 2.877,
      "step": 889
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.44205986059890934,
      "learning_rate": 0.0006699627929346227,
      "loss": 2.9355,
      "step": 890
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.44780292695690943,
      "learning_rate": 0.0006692641380431879,
      "loss": 3.041,
      "step": 891
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4170942494994364,
      "learning_rate": 0.0006685651096919393,
      "loss": 2.8887,
      "step": 892
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.40770218938046116,
      "learning_rate": 0.0006678657094231944,
      "loss": 2.959,
      "step": 893
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.49873111109204826,
      "learning_rate": 0.0006671659387800909,
      "loss": 2.9727,
      "step": 894
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4442025027835409,
      "learning_rate": 0.000666465799306584,
      "loss": 2.9551,
      "step": 895
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.43425074715676276,
      "learning_rate": 0.0006657652925474423,
      "loss": 2.9707,
      "step": 896
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4302927831256586,
      "learning_rate": 0.000665064420048245,
      "loss": 2.9473,
      "step": 897
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.49769405605332007,
      "learning_rate": 0.0006643631833553785,
      "loss": 2.9863,
      "step": 898
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4616689152141774,
      "learning_rate": 0.000663661584016032,
      "loss": 2.9629,
      "step": 899
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.5076193295955961,
      "learning_rate": 0.0006629596235781957,
      "loss": 2.9707,
      "step": 900
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4505022620267103,
      "learning_rate": 0.0006622573035906556,
      "loss": 2.9902,
      "step": 901
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.482290809024154,
      "learning_rate": 0.0006615546256029921,
      "loss": 2.8672,
      "step": 902
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.38675399035551167,
      "learning_rate": 0.0006608515911655743,
      "loss": 2.9414,
      "step": 903
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.43217870370126704,
      "learning_rate": 0.0006601482018295591,
      "loss": 2.9434,
      "step": 904
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4234115326148562,
      "learning_rate": 0.0006594444591468851,
      "loss": 2.9746,
      "step": 905
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4347798288440163,
      "learning_rate": 0.0006587403646702713,
      "loss": 2.9609,
      "step": 906
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.39568495512407026,
      "learning_rate": 0.0006580359199532126,
      "loss": 2.9238,
      "step": 907
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4578959659781502,
      "learning_rate": 0.000657331126549977,
      "loss": 2.8848,
      "step": 908
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.429220311630607,
      "learning_rate": 0.0006566259860156014,
      "loss": 2.8828,
      "step": 909
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.513047570542121,
      "learning_rate": 0.0006559204999058888,
      "loss": 2.9375,
      "step": 910
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4291318651327471,
      "learning_rate": 0.0006552146697774049,
      "loss": 2.9023,
      "step": 911
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4758472971946922,
      "learning_rate": 0.0006545084971874737,
      "loss": 2.9023,
      "step": 912
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4897740246527797,
      "learning_rate": 0.0006538019836941758,
      "loss": 2.9355,
      "step": 913
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.48491596868326736,
      "learning_rate": 0.0006530951308563431,
      "loss": 2.9648,
      "step": 914
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4486361602056078,
      "learning_rate": 0.0006523879402335567,
      "loss": 2.959,
      "step": 915
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4730875448243898,
      "learning_rate": 0.0006516804133861429,
      "loss": 3.084,
      "step": 916
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.5086387006361094,
      "learning_rate": 0.0006509725518751698,
      "loss": 2.8809,
      "step": 917
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.45749891514827434,
      "learning_rate": 0.0006502643572624438,
      "loss": 2.8945,
      "step": 918
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4155658310891238,
      "learning_rate": 0.0006495558311105064,
      "loss": 2.9121,
      "step": 919
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4324870319092521,
      "learning_rate": 0.0006488469749826305,
      "loss": 2.9473,
      "step": 920
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.43820714190283466,
      "learning_rate": 0.000648137790442817,
      "loss": 2.9277,
      "step": 921
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4780784396123727,
      "learning_rate": 0.0006474282790557916,
      "loss": 3.0195,
      "step": 922
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4904806152230454,
      "learning_rate": 0.000646718442387001,
      "loss": 3.0176,
      "step": 923
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4869929580472208,
      "learning_rate": 0.0006460082820026094,
      "loss": 2.9355,
      "step": 924
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4617023722328096,
      "learning_rate": 0.0006452977994694959,
      "loss": 2.9805,
      "step": 925
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.43940719321404115,
      "learning_rate": 0.0006445869963552496,
      "loss": 2.9668,
      "step": 926
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4760779654220946,
      "learning_rate": 0.0006438758742281672,
      "loss": 3.0215,
      "step": 927
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.45151500847253434,
      "learning_rate": 0.0006431644346572495,
      "loss": 2.9688,
      "step": 928
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.40892278446736685,
      "learning_rate": 0.0006424526792121974,
      "loss": 2.8867,
      "step": 929
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.43428150368660984,
      "learning_rate": 0.0006417406094634089,
      "loss": 2.9238,
      "step": 930
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.45410164543376696,
      "learning_rate": 0.0006410282269819756,
      "loss": 3.0117,
      "step": 931
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.48384674971897335,
      "learning_rate": 0.0006403155333396787,
      "loss": 3.002,
      "step": 932
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4639387186323789,
      "learning_rate": 0.0006396025301089863,
      "loss": 3.0352,
      "step": 933
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4400790766981229,
      "learning_rate": 0.0006388892188630493,
      "loss": 3.0312,
      "step": 934
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.44182640671526185,
      "learning_rate": 0.0006381756011756982,
      "loss": 3.0371,
      "step": 935
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4417290078873283,
      "learning_rate": 0.0006374616786214403,
      "loss": 2.9375,
      "step": 936
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.45954773575278135,
      "learning_rate": 0.0006367474527754544,
      "loss": 2.9629,
      "step": 937
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.48230564231689993,
      "learning_rate": 0.0006360329252135894,
      "loss": 2.9277,
      "step": 938
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.45912058326584126,
      "learning_rate": 0.0006353180975123595,
      "loss": 2.9688,
      "step": 939
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4362566194033281,
      "learning_rate": 0.0006346029712489413,
      "loss": 2.9648,
      "step": 940
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4316600133688967,
      "learning_rate": 0.0006338875480011698,
      "loss": 2.9746,
      "step": 941
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4581141332889205,
      "learning_rate": 0.0006331718293475357,
      "loss": 2.8867,
      "step": 942
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4522433511210253,
      "learning_rate": 0.0006324558168671811,
      "loss": 3.1133,
      "step": 943
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4397455055088126,
      "learning_rate": 0.0006317395121398968,
      "loss": 3.0156,
      "step": 944
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.5363739751091554,
      "learning_rate": 0.0006310229167461179,
      "loss": 2.9355,
      "step": 945
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.5272432427966016,
      "learning_rate": 0.0006303060322669214,
      "loss": 2.9629,
      "step": 946
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4337459124844383,
      "learning_rate": 0.0006295888602840214,
      "loss": 2.8301,
      "step": 947
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.47742648101244106,
      "learning_rate": 0.0006288714023797671,
      "loss": 2.959,
      "step": 948
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.467546279626172,
      "learning_rate": 0.000628153660137138,
      "loss": 2.918,
      "step": 949
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.45059321778295874,
      "learning_rate": 0.0006274356351397413,
      "loss": 2.9238,
      "step": 950
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3931062636322022,
      "learning_rate": 0.0006267173289718079,
      "loss": 2.9004,
      "step": 951
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4335257188148465,
      "learning_rate": 0.000625998743218189,
      "loss": 3.0059,
      "step": 952
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.45149929916348625,
      "learning_rate": 0.000625279879464353,
      "loss": 2.9062,
      "step": 953
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4680390787086187,
      "learning_rate": 0.000624560739296381,
      "loss": 2.9785,
      "step": 954
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4209400989811037,
      "learning_rate": 0.0006238413243009648,
      "loss": 2.8789,
      "step": 955
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.43237820017235234,
      "learning_rate": 0.000623121636065402,
      "loss": 2.8945,
      "step": 956
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4802314303160503,
      "learning_rate": 0.0006224016761775933,
      "loss": 2.9531,
      "step": 957
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4493190168869928,
      "learning_rate": 0.0006216814462260386,
      "loss": 2.9727,
      "step": 958
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.44005357364776293,
      "learning_rate": 0.0006209609477998338,
      "loss": 2.9785,
      "step": 959
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.42301782959237666,
      "learning_rate": 0.0006202401824886674,
      "loss": 2.9512,
      "step": 960
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.45835426042876437,
      "learning_rate": 0.0006195191518828162,
      "loss": 2.9668,
      "step": 961
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.47810193609600155,
      "learning_rate": 0.0006187978575731427,
      "loss": 2.9648,
      "step": 962
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4488630440110083,
      "learning_rate": 0.0006180763011510911,
      "loss": 2.9492,
      "step": 963
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4290939680129228,
      "learning_rate": 0.000617354484208684,
      "loss": 2.8906,
      "step": 964
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4229014507439965,
      "learning_rate": 0.0006166324083385189,
      "loss": 2.9648,
      "step": 965
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.400408161574504,
      "learning_rate": 0.0006159100751337642,
      "loss": 2.8867,
      "step": 966
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4460328035999983,
      "learning_rate": 0.0006151874861881565,
      "loss": 2.959,
      "step": 967
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4187959382008183,
      "learning_rate": 0.0006144646430959964,
      "loss": 2.9922,
      "step": 968
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4284905791814578,
      "learning_rate": 0.0006137415474521454,
      "loss": 2.9629,
      "step": 969
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.4721976547514943,
      "learning_rate": 0.0006130182008520222,
      "loss": 2.9414,
      "step": 970
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4204868350286959,
      "learning_rate": 0.000612294604891599,
      "loss": 2.9648,
      "step": 971
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4650516417349393,
      "learning_rate": 0.0006115707611673986,
      "loss": 2.9805,
      "step": 972
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4486045069752802,
      "learning_rate": 0.0006108466712764902,
      "loss": 2.918,
      "step": 973
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.46170258885567295,
      "learning_rate": 0.0006101223368164858,
      "loss": 3.0098,
      "step": 974
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.49810862868347994,
      "learning_rate": 0.0006093977593855375,
      "loss": 2.9336,
      "step": 975
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.5273170652911988,
      "learning_rate": 0.0006086729405823335,
      "loss": 2.957,
      "step": 976
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.482316397212794,
      "learning_rate": 0.0006079478820060943,
      "loss": 2.957,
      "step": 977
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.48687030287391114,
      "learning_rate": 0.0006072225852565695,
      "loss": 2.918,
      "step": 978
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.428744711671102,
      "learning_rate": 0.0006064970519340341,
      "loss": 2.9551,
      "step": 979
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4626250579748986,
      "learning_rate": 0.0006057712836392856,
      "loss": 2.9473,
      "step": 980
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4216422580656467,
      "learning_rate": 0.0006050452819736389,
      "loss": 2.9434,
      "step": 981
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.44798878512696355,
      "learning_rate": 0.000604319048538925,
      "loss": 2.9004,
      "step": 982
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.46648678083870315,
      "learning_rate": 0.0006035925849374855,
      "loss": 2.9824,
      "step": 983
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4718361249228428,
      "learning_rate": 0.0006028658927721697,
      "loss": 2.9844,
      "step": 984
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4265473769186804,
      "learning_rate": 0.0006021389736463321,
      "loss": 2.9805,
      "step": 985
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.433021842281239,
      "learning_rate": 0.0006014118291638271,
      "loss": 2.8809,
      "step": 986
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4778005243106488,
      "learning_rate": 0.0006006844609290065,
      "loss": 3.0586,
      "step": 987
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.47582455935734835,
      "learning_rate": 0.0005999568705467161,
      "loss": 3.084,
      "step": 988
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.43001081016913445,
      "learning_rate": 0.0005992290596222915,
      "loss": 2.9082,
      "step": 989
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4442722873634369,
      "learning_rate": 0.0005985010297615551,
      "loss": 2.9902,
      "step": 990
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.48460891560306035,
      "learning_rate": 0.0005977727825708123,
      "loss": 2.9785,
      "step": 991
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.5124763943838129,
      "learning_rate": 0.0005970443196568478,
      "loss": 2.9219,
      "step": 992
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4649023506323169,
      "learning_rate": 0.0005963156426269227,
      "loss": 2.8828,
      "step": 993
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4803914199251836,
      "learning_rate": 0.0005955867530887702,
      "loss": 2.9531,
      "step": 994
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.44324320595126626,
      "learning_rate": 0.0005948576526505923,
      "loss": 2.9219,
      "step": 995
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.44956043082881364,
      "learning_rate": 0.0005941283429210568,
      "loss": 2.9023,
      "step": 996
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4308341268594152,
      "learning_rate": 0.0005933988255092926,
      "loss": 2.9668,
      "step": 997
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.46186718724311093,
      "learning_rate": 0.0005926691020248874,
      "loss": 2.9141,
      "step": 998
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4620083768121919,
      "learning_rate": 0.0005919391740778833,
      "loss": 2.9531,
      "step": 999
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.44009340157391913,
      "learning_rate": 0.0005912090432787736,
      "loss": 2.9551,
      "step": 1000
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4213968619156871,
      "learning_rate": 0.000590478711238499,
      "loss": 2.8477,
      "step": 1001
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4371964869213959,
      "learning_rate": 0.0005897481795684446,
      "loss": 2.957,
      "step": 1002
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.45560592784540394,
      "learning_rate": 0.0005890174498804355,
      "loss": 2.8926,
      "step": 1003
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.46344724946054455,
      "learning_rate": 0.0005882865237867339,
      "loss": 2.9648,
      "step": 1004
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4615571926398025,
      "learning_rate": 0.0005875554029000353,
      "loss": 2.9355,
      "step": 1005
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4405406721669995,
      "learning_rate": 0.0005868240888334653,
      "loss": 2.8477,
      "step": 1006
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.46368076910899425,
      "learning_rate": 0.0005860925832005753,
      "loss": 2.8711,
      "step": 1007
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5098238535853685,
      "learning_rate": 0.0005853608876153395,
      "loss": 3.0137,
      "step": 1008
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5380301703542982,
      "learning_rate": 0.0005846290036921512,
      "loss": 2.9531,
      "step": 1009
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.49424270081042126,
      "learning_rate": 0.0005838969330458195,
      "loss": 2.834,
      "step": 1010
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.533370470328594,
      "learning_rate": 0.0005831646772915651,
      "loss": 2.9961,
      "step": 1011
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4463067655109399,
      "learning_rate": 0.0005824322380450173,
      "loss": 2.8789,
      "step": 1012
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4321884581647207,
      "learning_rate": 0.0005816996169222102,
      "loss": 3.0,
      "step": 1013
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4520467764323603,
      "learning_rate": 0.0005809668155395793,
      "loss": 2.9062,
      "step": 1014
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4568109959091713,
      "learning_rate": 0.0005802338355139578,
      "loss": 2.9316,
      "step": 1015
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.464275389460408,
      "learning_rate": 0.0005795006784625728,
      "loss": 2.9219,
      "step": 1016
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.44466826173405166,
      "learning_rate": 0.0005787673460030423,
      "loss": 2.9883,
      "step": 1017
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4266048352666604,
      "learning_rate": 0.000578033839753371,
      "loss": 2.8711,
      "step": 1018
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.44429039829851924,
      "learning_rate": 0.0005773001613319476,
      "loss": 2.9043,
      "step": 1019
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4834128985445496,
      "learning_rate": 0.00057656631235754,
      "loss": 2.8965,
      "step": 1020
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4953703977303074,
      "learning_rate": 0.0005758322944492929,
      "loss": 2.8906,
      "step": 1021
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.46167955957123774,
      "learning_rate": 0.0005750981092267237,
      "loss": 3.0273,
      "step": 1022
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4582095030934102,
      "learning_rate": 0.0005743637583097183,
      "loss": 2.8926,
      "step": 1023
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.5212103138271432,
      "learning_rate": 0.0005736292433185291,
      "loss": 2.9629,
      "step": 1024
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.46563722584472245,
      "learning_rate": 0.0005728945658737699,
      "loss": 3.0332,
      "step": 1025
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4644023150736749,
      "learning_rate": 0.0005721597275964133,
      "loss": 2.916,
      "step": 1026
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.42508092444233014,
      "learning_rate": 0.0005714247301077865,
      "loss": 2.8984,
      "step": 1027
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.43746651710673484,
      "learning_rate": 0.0005706895750295682,
      "loss": 2.9043,
      "step": 1028
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4470581878707031,
      "learning_rate": 0.0005699542639837844,
      "loss": 2.9453,
      "step": 1029
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.46726002745192746,
      "learning_rate": 0.0005692187985928055,
      "loss": 2.9336,
      "step": 1030
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.46083954624833123,
      "learning_rate": 0.0005684831804793427,
      "loss": 2.9414,
      "step": 1031
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.466459745729961,
      "learning_rate": 0.0005677474112664438,
      "loss": 3.0039,
      "step": 1032
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4988932845163524,
      "learning_rate": 0.0005670114925774899,
      "loss": 2.8965,
      "step": 1033
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4659327371289475,
      "learning_rate": 0.0005662754260361924,
      "loss": 3.0156,
      "step": 1034
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4173198286670981,
      "learning_rate": 0.0005655392132665884,
      "loss": 2.9336,
      "step": 1035
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.4585888068596422,
      "learning_rate": 0.000564802855893038,
      "loss": 2.9414,
      "step": 1036
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.4765698159559632,
      "learning_rate": 0.0005640663555402198,
      "loss": 2.9316,
      "step": 1037
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.46376231118490235,
      "learning_rate": 0.0005633297138331285,
      "loss": 2.8574,
      "step": 1038
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.5185784802640311,
      "learning_rate": 0.0005625929323970705,
      "loss": 2.9199,
      "step": 1039
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6041018731146975,
      "learning_rate": 0.0005618560128576603,
      "loss": 2.959,
      "step": 1040
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.45156245260631295,
      "learning_rate": 0.0005611189568408173,
      "loss": 2.8926,
      "step": 1041
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.5679654364783555,
      "learning_rate": 0.0005603817659727619,
      "loss": 2.9512,
      "step": 1042
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.4957712049903975,
      "learning_rate": 0.0005596444418800121,
      "loss": 2.9707,
      "step": 1043
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.434217597305553,
      "learning_rate": 0.0005589069861893798,
      "loss": 2.8848,
      "step": 1044
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.4668375096955166,
      "learning_rate": 0.0005581694005279673,
      "loss": 2.9551,
      "step": 1045
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.46857189328224336,
      "learning_rate": 0.0005574316865231637,
      "loss": 3.0508,
      "step": 1046
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.5129201615104234,
      "learning_rate": 0.0005566938458026411,
      "loss": 2.9492,
      "step": 1047
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.45864834224818585,
      "learning_rate": 0.0005559558799943514,
      "loss": 2.9707,
      "step": 1048
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.522315088808679,
      "learning_rate": 0.0005552177907265223,
      "loss": 2.9609,
      "step": 1049
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.4872022111745163,
      "learning_rate": 0.000554479579627654,
      "loss": 2.9043,
      "step": 1050
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.44442319239250283,
      "learning_rate": 0.0005537412483265157,
      "loss": 2.9844,
      "step": 1051
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.4707734314658287,
      "learning_rate": 0.0005530027984521413,
      "loss": 2.9824,
      "step": 1052
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.47619413947060524,
      "learning_rate": 0.0005522642316338268,
      "loss": 2.9824,
      "step": 1053
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3993280662919674,
      "learning_rate": 0.0005515255495011259,
      "loss": 2.8887,
      "step": 1054
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.46321712180425983,
      "learning_rate": 0.0005507867536838472,
      "loss": 2.9336,
      "step": 1055
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.46283916538061715,
      "learning_rate": 0.0005500478458120492,
      "loss": 2.9629,
      "step": 1056
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.4903848800731272,
      "learning_rate": 0.0005493088275160387,
      "loss": 3.0137,
      "step": 1057
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.44663836295874926,
      "learning_rate": 0.0005485697004263657,
      "loss": 2.9395,
      "step": 1058
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.39588234436431147,
      "learning_rate": 0.0005478304661738199,
      "loss": 2.9727,
      "step": 1059
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5164798943883854,
      "learning_rate": 0.0005470911263894279,
      "loss": 2.832,
      "step": 1060
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5453929793296344,
      "learning_rate": 0.0005463516827044491,
      "loss": 2.8594,
      "step": 1061
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5122110792360092,
      "learning_rate": 0.000545612136750372,
      "loss": 2.9648,
      "step": 1062
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.43378692347966585,
      "learning_rate": 0.0005448724901589107,
      "loss": 2.9492,
      "step": 1063
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.47420957491156296,
      "learning_rate": 0.0005441327445620014,
      "loss": 2.8965,
      "step": 1064
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.4934660230849155,
      "learning_rate": 0.0005433929015917988,
      "loss": 2.9219,
      "step": 1065
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.4516576758844985,
      "learning_rate": 0.0005426529628806724,
      "loss": 2.8516,
      "step": 1066
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5147054810732793,
      "learning_rate": 0.0005419129300612029,
      "loss": 2.8926,
      "step": 1067
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.4676630298864944,
      "learning_rate": 0.000541172804766179,
      "loss": 2.9629,
      "step": 1068
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.47547409486727127,
      "learning_rate": 0.0005404325886285927,
      "loss": 2.877,
      "step": 1069
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.5107692648418558,
      "learning_rate": 0.000539692283281637,
      "loss": 2.9766,
      "step": 1070
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.46738559902502685,
      "learning_rate": 0.0005389518903587017,
      "loss": 2.9883,
      "step": 1071
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.45992005040814415,
      "learning_rate": 0.0005382114114933695,
      "loss": 2.9043,
      "step": 1072
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.4575987144800551,
      "learning_rate": 0.0005374708483194132,
      "loss": 2.8711,
      "step": 1073
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.48573652668304323,
      "learning_rate": 0.000536730202470791,
      "loss": 2.9238,
      "step": 1074
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.43841653897023897,
      "learning_rate": 0.0005359894755816443,
      "loss": 2.916,
      "step": 1075
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.48514315985756257,
      "learning_rate": 0.0005352486692862926,
      "loss": 2.9023,
      "step": 1076
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.4431383754793275,
      "learning_rate": 0.0005345077852192307,
      "loss": 2.9375,
      "step": 1077
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.412603209249384,
      "learning_rate": 0.0005337668250151254,
      "loss": 2.8809,
      "step": 1078
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.48616847659965623,
      "learning_rate": 0.0005330257903088111,
      "loss": 2.8418,
      "step": 1079
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.44757388857243635,
      "learning_rate": 0.000532284682735287,
      "loss": 2.959,
      "step": 1080
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.45819108518249585,
      "learning_rate": 0.0005315435039297124,
      "loss": 2.8965,
      "step": 1081
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.44921277048054853,
      "learning_rate": 0.0005308022555274046,
      "loss": 2.9492,
      "step": 1082
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.502781789870664,
      "learning_rate": 0.0005300609391638336,
      "loss": 3.0039,
      "step": 1083
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4220192798635597,
      "learning_rate": 0.0005293195564746201,
      "loss": 2.9082,
      "step": 1084
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.44507635338792506,
      "learning_rate": 0.0005285781090955304,
      "loss": 2.9434,
      "step": 1085
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.406022305264462,
      "learning_rate": 0.0005278365986624743,
      "loss": 2.875,
      "step": 1086
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.42053954605299015,
      "learning_rate": 0.0005270950268115001,
      "loss": 2.9824,
      "step": 1087
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.49176425421461833,
      "learning_rate": 0.0005263533951787919,
      "loss": 2.918,
      "step": 1088
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.46589227573867104,
      "learning_rate": 0.000525611705400666,
      "loss": 3.002,
      "step": 1089
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.43791642089250793,
      "learning_rate": 0.0005248699591135664,
      "loss": 2.9961,
      "step": 1090
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.43002476626601155,
      "learning_rate": 0.0005241281579540618,
      "loss": 2.8828,
      "step": 1091
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4784906076075858,
      "learning_rate": 0.0005233863035588427,
      "loss": 2.9922,
      "step": 1092
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4330278577246521,
      "learning_rate": 0.0005226443975647161,
      "loss": 2.9629,
      "step": 1093
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4585680421781931,
      "learning_rate": 0.0005219024416086036,
      "loss": 2.9082,
      "step": 1094
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4642599001348248,
      "learning_rate": 0.0005211604373275366,
      "loss": 2.9941,
      "step": 1095
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.45885764808072993,
      "learning_rate": 0.0005204183863586533,
      "loss": 2.9531,
      "step": 1096
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.41854865014704656,
      "learning_rate": 0.0005196762903391951,
      "loss": 2.959,
      "step": 1097
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4577258141121951,
      "learning_rate": 0.0005189341509065023,
      "loss": 2.8789,
      "step": 1098
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4413166034890246,
      "learning_rate": 0.0005181919696980112,
      "loss": 2.9297,
      "step": 1099
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4705529758410707,
      "learning_rate": 0.0005174497483512506,
      "loss": 2.9004,
      "step": 1100
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4210398245970409,
      "learning_rate": 0.0005167074885038374,
      "loss": 2.8809,
      "step": 1101
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4284940594778489,
      "learning_rate": 0.0005159651917934735,
      "loss": 2.8242,
      "step": 1102
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4584239431569254,
      "learning_rate": 0.0005152228598579428,
      "loss": 2.9355,
      "step": 1103
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.44471899776072876,
      "learning_rate": 0.000514480494335106,
      "loss": 2.875,
      "step": 1104
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4334910142577977,
      "learning_rate": 0.0005137380968628983,
      "loss": 2.9785,
      "step": 1105
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.44630031055032204,
      "learning_rate": 0.0005129956690793255,
      "loss": 2.9297,
      "step": 1106
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4869749440948907,
      "learning_rate": 0.0005122532126224601,
      "loss": 2.8984,
      "step": 1107
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4525111285500587,
      "learning_rate": 0.0005115107291304378,
      "loss": 2.8945,
      "step": 1108
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.44412978559270394,
      "learning_rate": 0.0005107682202414544,
      "loss": 2.9473,
      "step": 1109
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4090851289552635,
      "learning_rate": 0.0005100256875937613,
      "loss": 2.9316,
      "step": 1110
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4346919928463004,
      "learning_rate": 0.0005092831328256625,
      "loss": 2.9492,
      "step": 1111
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.5051308433141847,
      "learning_rate": 0.0005085405575755105,
      "loss": 2.9668,
      "step": 1112
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.44177145903811826,
      "learning_rate": 0.0005077979634817034,
      "loss": 2.9434,
      "step": 1113
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4502233905838624,
      "learning_rate": 0.0005070553521826808,
      "loss": 2.9805,
      "step": 1114
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4327074020346755,
      "learning_rate": 0.00050631272531692,
      "loss": 2.8691,
      "step": 1115
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4962492498994229,
      "learning_rate": 0.0005055700845229327,
      "loss": 2.9824,
      "step": 1116
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4446376274799597,
      "learning_rate": 0.000504827431439262,
      "loss": 2.9668,
      "step": 1117
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.45514278978248723,
      "learning_rate": 0.000504084767704477,
      "loss": 2.873,
      "step": 1118
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4590667416090236,
      "learning_rate": 0.0005033420949571712,
      "loss": 3.0078,
      "step": 1119
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.48530907287228314,
      "learning_rate": 0.0005025994148359574,
      "loss": 2.8379,
      "step": 1120
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4853337619719855,
      "learning_rate": 0.0005018567289794651,
      "loss": 2.8848,
      "step": 1121
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.5261083145302712,
      "learning_rate": 0.0005011140390263362,
      "loss": 2.9707,
      "step": 1122
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.46257039418845997,
      "learning_rate": 0.0005003713466152218,
      "loss": 2.8574,
      "step": 1123
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.41053177252011536,
      "learning_rate": 0.0004996286533847783,
      "loss": 2.9102,
      "step": 1124
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.44492033567467915,
      "learning_rate": 0.000498885960973664,
      "loss": 2.8965,
      "step": 1125
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4288384279833731,
      "learning_rate": 0.000498143271020535,
      "loss": 2.9297,
      "step": 1126
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4343445042217144,
      "learning_rate": 0.0004974005851640428,
      "loss": 2.9805,
      "step": 1127
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.43623867700360536,
      "learning_rate": 0.000496657905042829,
      "loss": 3.0195,
      "step": 1128
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4203803067972236,
      "learning_rate": 0.0004959152322955232,
      "loss": 2.9707,
      "step": 1129
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4911744442375095,
      "learning_rate": 0.0004951725685607382,
      "loss": 2.8496,
      "step": 1130
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.42284976072896624,
      "learning_rate": 0.0004944299154770673,
      "loss": 2.8301,
      "step": 1131
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4714628902607225,
      "learning_rate": 0.0004936872746830802,
      "loss": 2.8926,
      "step": 1132
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4798112177756236,
      "learning_rate": 0.0004929446478173195,
      "loss": 2.8711,
      "step": 1133
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4883262277391658,
      "learning_rate": 0.0004922020365182968,
      "loss": 2.9492,
      "step": 1134
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5127293133814586,
      "learning_rate": 0.0004914594424244897,
      "loss": 2.9453,
      "step": 1135
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4757867751998179,
      "learning_rate": 0.0004907168671743376,
      "loss": 2.9082,
      "step": 1136
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.49022387737574663,
      "learning_rate": 0.0004899743124062387,
      "loss": 2.998,
      "step": 1137
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5076262605083499,
      "learning_rate": 0.0004892317797585456,
      "loss": 2.9492,
      "step": 1138
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.46075408846220545,
      "learning_rate": 0.0004884892708695623,
      "loss": 2.8984,
      "step": 1139
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5019241355772536,
      "learning_rate": 0.0004877467873775402,
      "loss": 2.8926,
      "step": 1140
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.49031328890867426,
      "learning_rate": 0.00048700433092067473,
      "loss": 2.9355,
      "step": 1141
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5120133211461211,
      "learning_rate": 0.0004862619031371019,
      "loss": 2.9707,
      "step": 1142
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4732618321829559,
      "learning_rate": 0.0004855195056648942,
      "loss": 2.9355,
      "step": 1143
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4588399360207129,
      "learning_rate": 0.00048477714014205734,
      "loss": 2.8379,
      "step": 1144
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.4973588910986723,
      "learning_rate": 0.00048403480820652644,
      "loss": 2.9062,
      "step": 1145
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5065342637793789,
      "learning_rate": 0.0004832925114961629,
      "loss": 2.9805,
      "step": 1146
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4862406120917773,
      "learning_rate": 0.0004825502516487497,
      "loss": 2.8457,
      "step": 1147
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5049195747473509,
      "learning_rate": 0.00048180803030198896,
      "loss": 3.0176,
      "step": 1148
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5049831636058524,
      "learning_rate": 0.0004810658490934979,
      "loss": 2.9922,
      "step": 1149
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5150653422429855,
      "learning_rate": 0.000480323709660805,
      "loss": 2.8594,
      "step": 1150
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.43865069933112194,
      "learning_rate": 0.0004795816136413467,
      "loss": 2.9219,
      "step": 1151
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.43603062096706235,
      "learning_rate": 0.00047883956267246353,
      "loss": 2.8789,
      "step": 1152
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4418010211634295,
      "learning_rate": 0.00047809755839139657,
      "loss": 2.8555,
      "step": 1153
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4938587742616279,
      "learning_rate": 0.0004773556024352841,
      "loss": 2.9863,
      "step": 1154
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4386877358409651,
      "learning_rate": 0.00047661369644115754,
      "loss": 3.0176,
      "step": 1155
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4943659197194051,
      "learning_rate": 0.0004758718420459383,
      "loss": 2.918,
      "step": 1156
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4535985313307704,
      "learning_rate": 0.0004751300408864339,
      "loss": 2.998,
      "step": 1157
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4296960104091406,
      "learning_rate": 0.00047438829459933414,
      "loss": 2.8984,
      "step": 1158
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.45037475369388724,
      "learning_rate": 0.0004736466048212082,
      "loss": 2.9355,
      "step": 1159
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4531590207714843,
      "learning_rate": 0.0004729049731885002,
      "loss": 2.8945,
      "step": 1160
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5326392932477616,
      "learning_rate": 0.000472163401337526,
      "loss": 2.8145,
      "step": 1161
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.40762192754538157,
      "learning_rate": 0.00047142189090446985,
      "loss": 2.9316,
      "step": 1162
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.46704900572597874,
      "learning_rate": 0.0004706804435253802,
      "loss": 2.8008,
      "step": 1163
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.46972583274526875,
      "learning_rate": 0.0004699390608361665,
      "loss": 2.9297,
      "step": 1164
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.45893424998796295,
      "learning_rate": 0.0004691977444725955,
      "loss": 2.8789,
      "step": 1165
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.4433139455307656,
      "learning_rate": 0.0004684564960702877,
      "loss": 2.9648,
      "step": 1166
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.48405563682776925,
      "learning_rate": 0.0004677153172647131,
      "loss": 2.9434,
      "step": 1167
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.5072611965197439,
      "learning_rate": 0.00046697420969118894,
      "loss": 2.8926,
      "step": 1168
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4932204262703781,
      "learning_rate": 0.00046623317498487466,
      "loss": 2.8301,
      "step": 1169
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.5127411673396239,
      "learning_rate": 0.0004654922147807694,
      "loss": 2.8887,
      "step": 1170
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4891713153914133,
      "learning_rate": 0.00046475133071370757,
      "loss": 2.8379,
      "step": 1171
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4748017029700832,
      "learning_rate": 0.00046401052441835574,
      "loss": 2.9531,
      "step": 1172
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.48862993530909316,
      "learning_rate": 0.000463269797529209,
      "loss": 2.8887,
      "step": 1173
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.43858008905551593,
      "learning_rate": 0.00046252915168058697,
      "loss": 2.8906,
      "step": 1174
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.43058203072187096,
      "learning_rate": 0.0004617885885066305,
      "loss": 2.916,
      "step": 1175
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.46061333446828434,
      "learning_rate": 0.0004610481096412984,
      "loss": 2.9492,
      "step": 1176
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.5149163093836383,
      "learning_rate": 0.000460307716718363,
      "loss": 2.957,
      "step": 1177
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.49204445136266556,
      "learning_rate": 0.0004595674113714074,
      "loss": 2.9336,
      "step": 1178
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.47560365104023444,
      "learning_rate": 0.0004588271952338212,
      "loss": 2.8496,
      "step": 1179
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4974663572199639,
      "learning_rate": 0.00045808706993879714,
      "loss": 2.8555,
      "step": 1180
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.5149654714524121,
      "learning_rate": 0.00045734703711932767,
      "loss": 2.918,
      "step": 1181
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4386582675766582,
      "learning_rate": 0.0004566070984082013,
      "loss": 3.0,
      "step": 1182
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.463335666087367,
      "learning_rate": 0.00045586725543799865,
      "loss": 2.9609,
      "step": 1183
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4597742196575045,
      "learning_rate": 0.00045512750984108937,
      "loss": 2.9746,
      "step": 1184
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.46947513883141967,
      "learning_rate": 0.000454387863249628,
      "loss": 2.9473,
      "step": 1185
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4667456729242712,
      "learning_rate": 0.00045364831729555096,
      "loss": 2.9473,
      "step": 1186
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4837268335503023,
      "learning_rate": 0.0004529088736105721,
      "loss": 3.002,
      "step": 1187
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4456141114402603,
      "learning_rate": 0.0004521695338261802,
      "loss": 2.9609,
      "step": 1188
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.49396689436248653,
      "learning_rate": 0.0004514302995736344,
      "loss": 2.9199,
      "step": 1189
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.47310381294380105,
      "learning_rate": 0.0004506911724839613,
      "loss": 2.9492,
      "step": 1190
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4604527484944589,
      "learning_rate": 0.0004499521541879508,
      "loss": 2.9336,
      "step": 1191
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4315522202021604,
      "learning_rate": 0.00044921324631615303,
      "loss": 2.9062,
      "step": 1192
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.444290023405673,
      "learning_rate": 0.0004484744504988742,
      "loss": 2.9199,
      "step": 1193
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4325452651202194,
      "learning_rate": 0.00044773576836617336,
      "loss": 2.9102,
      "step": 1194
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4494344026518817,
      "learning_rate": 0.0004469972015478588,
      "loss": 2.9297,
      "step": 1195
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4321089729867976,
      "learning_rate": 0.0004462587516734844,
      "loss": 2.8867,
      "step": 1196
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4305820188257074,
      "learning_rate": 0.00044552042037234596,
      "loss": 2.9766,
      "step": 1197
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4349116605806353,
      "learning_rate": 0.00044478220927347774,
      "loss": 2.9707,
      "step": 1198
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4279889584149693,
      "learning_rate": 0.00044404412000564875,
      "loss": 2.8672,
      "step": 1199
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4106171378559405,
      "learning_rate": 0.000443306154197359,
      "loss": 2.9355,
      "step": 1200
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4191231911287359,
      "learning_rate": 0.00044256831347683646,
      "loss": 2.9707,
      "step": 1201
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4448109600193207,
      "learning_rate": 0.0004418305994720328,
      "loss": 2.9453,
      "step": 1202
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4456070890345972,
      "learning_rate": 0.0004410930138106203,
      "loss": 2.873,
      "step": 1203
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.45366918911083604,
      "learning_rate": 0.000440355558119988,
      "loss": 2.9473,
      "step": 1204
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4346401499703426,
      "learning_rate": 0.00043961823402723814,
      "loss": 2.875,
      "step": 1205
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4926321949466293,
      "learning_rate": 0.0004388810431591829,
      "loss": 2.9102,
      "step": 1206
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.42897847514265075,
      "learning_rate": 0.0004381439871423398,
      "loss": 2.9434,
      "step": 1207
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4974473389812989,
      "learning_rate": 0.00043740706760292966,
      "loss": 2.9512,
      "step": 1208
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.45576664289082025,
      "learning_rate": 0.0004366702861668716,
      "loss": 2.9043,
      "step": 1209
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.4501564258686837,
      "learning_rate": 0.00043593364445978036,
      "loss": 2.8594,
      "step": 1210
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.4893080396839686,
      "learning_rate": 0.0004351971441069622,
      "loss": 2.9316,
      "step": 1211
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.440037820785081,
      "learning_rate": 0.0004344607867334116,
      "loss": 2.9395,
      "step": 1212
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.48301890675166553,
      "learning_rate": 0.00043372457396380766,
      "loss": 2.9355,
      "step": 1213
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.4887567956972505,
      "learning_rate": 0.00043298850742251013,
      "loss": 2.9492,
      "step": 1214
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.49743083324571297,
      "learning_rate": 0.0004322525887335563,
      "loss": 2.8984,
      "step": 1215
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.4685286703154868,
      "learning_rate": 0.00043151681952065734,
      "loss": 2.9121,
      "step": 1216
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.463743612116786,
      "learning_rate": 0.00043078120140719456,
      "loss": 2.9414,
      "step": 1217
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.47805740634642085,
      "learning_rate": 0.0004300457360162158,
      "loss": 2.8594,
      "step": 1218
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.43742500457613515,
      "learning_rate": 0.0004293104249704319,
      "loss": 2.8945,
      "step": 1219
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.44509003525492685,
      "learning_rate": 0.00042857526989221355,
      "loss": 2.793,
      "step": 1220
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.47121409030527756,
      "learning_rate": 0.00042784027240358674,
      "loss": 2.9355,
      "step": 1221
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.4207261188723065,
      "learning_rate": 0.0004271054341262301,
      "loss": 2.8965,
      "step": 1222
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.4329643819368615,
      "learning_rate": 0.000426370756681471,
      "loss": 2.9277,
      "step": 1223
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.46423792162905186,
      "learning_rate": 0.0004256362416902817,
      "loss": 2.9512,
      "step": 1224
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.48186239300684036,
      "learning_rate": 0.00042490189077327637,
      "loss": 2.8926,
      "step": 1225
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.42714986500841645,
      "learning_rate": 0.00042416770555070703,
      "loss": 2.9219,
      "step": 1226
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.47656927762893997,
      "learning_rate": 0.00042343368764246,
      "loss": 2.9551,
      "step": 1227
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.47034378898184,
      "learning_rate": 0.0004226998386680524,
      "loss": 2.9492,
      "step": 1228
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.47849353671183825,
      "learning_rate": 0.000421966160246629,
      "loss": 2.9629,
      "step": 1229
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.48436612327290457,
      "learning_rate": 0.00042123265399695783,
      "loss": 2.8574,
      "step": 1230
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.4991858138225296,
      "learning_rate": 0.0004204993215374273,
      "loss": 2.9062,
      "step": 1231
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.48395004574414274,
      "learning_rate": 0.00041976616448604226,
      "loss": 2.9785,
      "step": 1232
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4465201326681686,
      "learning_rate": 0.00041903318446042076,
      "loss": 2.9355,
      "step": 1233
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4694946154693628,
      "learning_rate": 0.00041830038307778984,
      "loss": 2.918,
      "step": 1234
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.5157719665646703,
      "learning_rate": 0.0004175677619549828,
      "loss": 2.9629,
      "step": 1235
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4727553115594631,
      "learning_rate": 0.000416835322708435,
      "loss": 2.8398,
      "step": 1236
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4305419265159187,
      "learning_rate": 0.00041610306695418056,
      "loss": 2.8574,
      "step": 1237
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.45175754312326016,
      "learning_rate": 0.0004153709963078488,
      "loss": 2.8652,
      "step": 1238
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4761401283355484,
      "learning_rate": 0.0004146391123846606,
      "loss": 3.0234,
      "step": 1239
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4389577265396903,
      "learning_rate": 0.0004139074167994249,
      "loss": 2.8809,
      "step": 1240
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.45829170160217453,
      "learning_rate": 0.00041317591116653486,
      "loss": 2.9219,
      "step": 1241
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.43682190629560064,
      "learning_rate": 0.0004124445970999648,
      "loss": 2.8711,
      "step": 1242
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.45718329650299894,
      "learning_rate": 0.00041171347621326627,
      "loss": 3.0,
      "step": 1243
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4641771241670593,
      "learning_rate": 0.00041098255011956465,
      "loss": 2.8242,
      "step": 1244
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.44390967795913944,
      "learning_rate": 0.00041025182043155547,
      "loss": 2.9863,
      "step": 1245
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.48016243376620077,
      "learning_rate": 0.000409521288761501,
      "loss": 2.8164,
      "step": 1246
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4834419492503505,
      "learning_rate": 0.00040879095672122646,
      "loss": 2.8398,
      "step": 1247
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4796993724891628,
      "learning_rate": 0.0004080608259221167,
      "loss": 2.9043,
      "step": 1248
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4699065798202022,
      "learning_rate": 0.0004073308979751126,
      "loss": 2.9492,
      "step": 1249
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4742403106714077,
      "learning_rate": 0.0004066011744907074,
      "loss": 2.8281,
      "step": 1250
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4189789665497103,
      "learning_rate": 0.00040587165707894326,
      "loss": 2.9609,
      "step": 1251
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4930542665185297,
      "learning_rate": 0.0004051423473494076,
      "loss": 2.8809,
      "step": 1252
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.5236992014983872,
      "learning_rate": 0.0004044132469112299,
      "loss": 2.8535,
      "step": 1253
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.48824501248837293,
      "learning_rate": 0.00040368435737307733,
      "loss": 2.9375,
      "step": 1254
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.5179427377494988,
      "learning_rate": 0.00040295568034315224,
      "loss": 3.0312,
      "step": 1255
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4200815901686863,
      "learning_rate": 0.0004022272174291878,
      "loss": 2.9414,
      "step": 1256
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.5601089131915152,
      "learning_rate": 0.0004014989702384449,
      "loss": 2.8047,
      "step": 1257
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4678770185622242,
      "learning_rate": 0.00040077094037770843,
      "loss": 2.8945,
      "step": 1258
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4617443312531526,
      "learning_rate": 0.0004000431294532838,
      "loss": 2.9512,
      "step": 1259
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.46319799106023246,
      "learning_rate": 0.0003993155390709935,
      "loss": 2.9316,
      "step": 1260
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4508984020301487,
      "learning_rate": 0.0003985881708361729,
      "loss": 2.9297,
      "step": 1261
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.44832686760508444,
      "learning_rate": 0.00039786102635366784,
      "loss": 2.9121,
      "step": 1262
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.43509204435005416,
      "learning_rate": 0.0003971341072278302,
      "loss": 2.8262,
      "step": 1263
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.484491539378372,
      "learning_rate": 0.00039640741506251457,
      "loss": 2.8691,
      "step": 1264
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4753525589274207,
      "learning_rate": 0.00039568095146107495,
      "loss": 3.0254,
      "step": 1265
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4323658550150046,
      "learning_rate": 0.00039495471802636096,
      "loss": 2.9141,
      "step": 1266
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.43993133288355385,
      "learning_rate": 0.0003942287163607145,
      "loss": 2.8672,
      "step": 1267
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4968332909324718,
      "learning_rate": 0.0003935029480659658,
      "loss": 2.8848,
      "step": 1268
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.517835425578818,
      "learning_rate": 0.00039277741474343054,
      "loss": 2.8809,
      "step": 1269
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4768366832883377,
      "learning_rate": 0.0003920521179939057,
      "loss": 2.8926,
      "step": 1270
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4933544382512723,
      "learning_rate": 0.00039132705941766644,
      "loss": 2.918,
      "step": 1271
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.41674526349989127,
      "learning_rate": 0.0003906022406144624,
      "loss": 2.8203,
      "step": 1272
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.457831323075177,
      "learning_rate": 0.0003898776631835143,
      "loss": 2.9023,
      "step": 1273
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4968546684338373,
      "learning_rate": 0.00038915332872350994,
      "loss": 2.8184,
      "step": 1274
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.4541349408357168,
      "learning_rate": 0.00038842923883260135,
      "loss": 2.8926,
      "step": 1275
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.44442830360394225,
      "learning_rate": 0.00038770539510840093,
      "loss": 2.9238,
      "step": 1276
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.43411323309309785,
      "learning_rate": 0.00038698179914797783,
      "loss": 2.8574,
      "step": 1277
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.4918426318258786,
      "learning_rate": 0.0003862584525478545,
      "loss": 2.9199,
      "step": 1278
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.47041090775668937,
      "learning_rate": 0.00038553535690400353,
      "loss": 2.8398,
      "step": 1279
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.4914215027150734,
      "learning_rate": 0.00038481251381184355,
      "loss": 2.8359,
      "step": 1280
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.45393880521779534,
      "learning_rate": 0.00038408992486623584,
      "loss": 2.8848,
      "step": 1281
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.43970183437505983,
      "learning_rate": 0.00038336759166148117,
      "loss": 2.916,
      "step": 1282
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.48053208187905755,
      "learning_rate": 0.0003826455157913159,
      "loss": 2.8848,
      "step": 1283
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.478364223192804,
      "learning_rate": 0.00038192369884890886,
      "loss": 2.8867,
      "step": 1284
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.49979909809733686,
      "learning_rate": 0.00038120214242685723,
      "loss": 2.9824,
      "step": 1285
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.4640091454223826,
      "learning_rate": 0.00038048084811718373,
      "loss": 2.8691,
      "step": 1286
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.484274101359453,
      "learning_rate": 0.0003797598175113327,
      "loss": 2.9609,
      "step": 1287
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.5285727294335588,
      "learning_rate": 0.0003790390522001662,
      "loss": 2.9512,
      "step": 1288
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.4422991623011609,
      "learning_rate": 0.0003783185537739615,
      "loss": 2.9414,
      "step": 1289
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.4663078382683172,
      "learning_rate": 0.00037759832382240697,
      "loss": 2.9102,
      "step": 1290
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.4975848163535184,
      "learning_rate": 0.00037687836393459826,
      "loss": 2.8887,
      "step": 1291
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.43134357659437184,
      "learning_rate": 0.0003761586756990354,
      "loss": 2.9727,
      "step": 1292
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.43242956356334256,
      "learning_rate": 0.0003754392607036191,
      "loss": 2.8926,
      "step": 1293
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.49789892700703514,
      "learning_rate": 0.0003747201205356472,
      "loss": 2.9492,
      "step": 1294
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.44372852811795543,
      "learning_rate": 0.0003740012567818111,
      "loss": 2.8906,
      "step": 1295
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.4426529500139389,
      "learning_rate": 0.0003732826710281922,
      "loss": 2.9297,
      "step": 1296
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.4712989584628926,
      "learning_rate": 0.0003725643648602588,
      "loss": 2.9277,
      "step": 1297
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4530463422107368,
      "learning_rate": 0.0003718463398628621,
      "loss": 2.8789,
      "step": 1298
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.46148521902107836,
      "learning_rate": 0.0003711285976202331,
      "loss": 2.9453,
      "step": 1299
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.44215005476747277,
      "learning_rate": 0.0003704111397159787,
      "loss": 2.9043,
      "step": 1300
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4323253752387821,
      "learning_rate": 0.0003696939677330788,
      "loss": 2.8477,
      "step": 1301
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.46291355046608884,
      "learning_rate": 0.00036897708325388213,
      "loss": 2.8848,
      "step": 1302
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.5039137415319804,
      "learning_rate": 0.0003682604878601034,
      "loss": 2.8555,
      "step": 1303
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4701887968200789,
      "learning_rate": 0.000367544183132819,
      "loss": 2.8516,
      "step": 1304
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4769809474060334,
      "learning_rate": 0.0003668281706524645,
      "loss": 2.832,
      "step": 1305
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.46363219783966225,
      "learning_rate": 0.0003661124519988304,
      "loss": 2.8672,
      "step": 1306
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4885690556440086,
      "learning_rate": 0.00036539702875105893,
      "loss": 2.9258,
      "step": 1307
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4789129887511452,
      "learning_rate": 0.0003646819024876406,
      "loss": 2.834,
      "step": 1308
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.5111583930858724,
      "learning_rate": 0.0003639670747864107,
      "loss": 2.9023,
      "step": 1309
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4639988547667179,
      "learning_rate": 0.00036325254722454584,
      "loss": 2.9316,
      "step": 1310
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.472594608743489,
      "learning_rate": 0.00036253832137855997,
      "loss": 2.8613,
      "step": 1311
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4570420152935334,
      "learning_rate": 0.00036182439882430183,
      "loss": 2.918,
      "step": 1312
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4572440141472346,
      "learning_rate": 0.00036111078113695096,
      "loss": 2.9258,
      "step": 1313
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.46425002095532314,
      "learning_rate": 0.0003603974698910139,
      "loss": 2.957,
      "step": 1314
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.47010948785528456,
      "learning_rate": 0.0003596844666603214,
      "loss": 2.9102,
      "step": 1315
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4294090238971709,
      "learning_rate": 0.0003589717730180245,
      "loss": 2.8633,
      "step": 1316
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.47836357311358685,
      "learning_rate": 0.00035825939053659117,
      "loss": 2.9121,
      "step": 1317
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4755437205141799,
      "learning_rate": 0.00035754732078780273,
      "loss": 2.9629,
      "step": 1318
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.46696540560854727,
      "learning_rate": 0.00035683556534275076,
      "loss": 2.9668,
      "step": 1319
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4558466924503936,
      "learning_rate": 0.00035612412577183303,
      "loss": 2.8594,
      "step": 1320
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.47008464695264107,
      "learning_rate": 0.00035541300364475063,
      "loss": 2.916,
      "step": 1321
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.45599018104410405,
      "learning_rate": 0.0003547022005305043,
      "loss": 2.8887,
      "step": 1322
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.44356567175332584,
      "learning_rate": 0.0003539917179973907,
      "loss": 2.9023,
      "step": 1323
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4304085292582338,
      "learning_rate": 0.00035328155761299917,
      "loss": 2.9082,
      "step": 1324
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4782323072144959,
      "learning_rate": 0.0003525717209442085,
      "loss": 2.9551,
      "step": 1325
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4836104785293679,
      "learning_rate": 0.00035186220955718306,
      "loss": 2.9102,
      "step": 1326
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4527480019878289,
      "learning_rate": 0.0003511530250173696,
      "loss": 2.9238,
      "step": 1327
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4608101997962914,
      "learning_rate": 0.00035044416888949364,
      "loss": 2.9277,
      "step": 1328
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.5098099614485229,
      "learning_rate": 0.0003497356427375562,
      "loss": 2.9766,
      "step": 1329
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4706718399926077,
      "learning_rate": 0.00034902744812483034,
      "loss": 2.9824,
      "step": 1330
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.47369814463935955,
      "learning_rate": 0.00034831958661385714,
      "loss": 2.957,
      "step": 1331
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.45262429409736377,
      "learning_rate": 0.0003476120597664434,
      "loss": 2.8633,
      "step": 1332
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4429804771414118,
      "learning_rate": 0.00034690486914365704,
      "loss": 2.9395,
      "step": 1333
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4788848631692888,
      "learning_rate": 0.00034619801630582435,
      "loss": 2.8594,
      "step": 1334
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.5286304815867169,
      "learning_rate": 0.00034549150281252633,
      "loss": 2.9902,
      "step": 1335
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4768567705927634,
      "learning_rate": 0.0003447853302225952,
      "loss": 2.8887,
      "step": 1336
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.45302448077529633,
      "learning_rate": 0.00034407950009411126,
      "loss": 2.9492,
      "step": 1337
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.44693375970354304,
      "learning_rate": 0.00034337401398439873,
      "loss": 2.8789,
      "step": 1338
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4852614098074016,
      "learning_rate": 0.00034266887345002305,
      "loss": 2.9258,
      "step": 1339
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.476038886250771,
      "learning_rate": 0.0003419640800467874,
      "loss": 2.8984,
      "step": 1340
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.47576329265713946,
      "learning_rate": 0.0003412596353297288,
      "loss": 2.8516,
      "step": 1341
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.4872523606516207,
      "learning_rate": 0.00034055554085311493,
      "loss": 3.0137,
      "step": 1342
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.45747542299073574,
      "learning_rate": 0.00033985179817044105,
      "loss": 2.957,
      "step": 1343
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5069348219796659,
      "learning_rate": 0.0003391484088344257,
      "loss": 2.918,
      "step": 1344
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.4306961785760296,
      "learning_rate": 0.00033844537439700807,
      "loss": 2.8125,
      "step": 1345
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.45093180586696696,
      "learning_rate": 0.00033774269640934445,
      "loss": 2.9141,
      "step": 1346
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.43747106328305385,
      "learning_rate": 0.0003370403764218045,
      "loss": 2.8867,
      "step": 1347
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.4301646201185612,
      "learning_rate": 0.000336338415983968,
      "loss": 2.8945,
      "step": 1348
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.43839434574939745,
      "learning_rate": 0.00033563681664462155,
      "loss": 2.8516,
      "step": 1349
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.44766374056382324,
      "learning_rate": 0.000334935579951755,
      "loss": 2.9766,
      "step": 1350
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.4381691780451791,
      "learning_rate": 0.0003342347074525578,
      "loss": 2.8379,
      "step": 1351
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.4600961541098975,
      "learning_rate": 0.0003335342006934161,
      "loss": 2.9121,
      "step": 1352
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.4591462954608774,
      "learning_rate": 0.00033283406121990914,
      "loss": 3.002,
      "step": 1353
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.43256620146752167,
      "learning_rate": 0.0003321342905768057,
      "loss": 2.9102,
      "step": 1354
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.44689670167660966,
      "learning_rate": 0.00033143489030806086,
      "loss": 2.8652,
      "step": 1355
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.4274714599633656,
      "learning_rate": 0.00033073586195681227,
      "loss": 2.8398,
      "step": 1356
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.4889159000385258,
      "learning_rate": 0.00033003720706537736,
      "loss": 2.8711,
      "step": 1357
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.539129933632698,
      "learning_rate": 0.0003293389271752492,
      "loss": 2.7871,
      "step": 1358
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.4596649018826197,
      "learning_rate": 0.00032864102382709374,
      "loss": 2.8359,
      "step": 1359
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5261205273764837,
      "learning_rate": 0.000327943498560746,
      "loss": 2.9414,
      "step": 1360
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5001544042269518,
      "learning_rate": 0.00032724635291520694,
      "loss": 2.875,
      "step": 1361
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5217395188555193,
      "learning_rate": 0.00032654958842863967,
      "loss": 2.8848,
      "step": 1362
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5734215212987728,
      "learning_rate": 0.0003258532066383667,
      "loss": 2.8457,
      "step": 1363
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.4706589629352283,
      "learning_rate": 0.000325157209080866,
      "loss": 2.8828,
      "step": 1364
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.48509364199733157,
      "learning_rate": 0.00032446159729176743,
      "loss": 2.9551,
      "step": 1365
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5086527963858644,
      "learning_rate": 0.0003237663728058502,
      "loss": 2.9082,
      "step": 1366
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.45264853854394416,
      "learning_rate": 0.0003230715371570389,
      "loss": 2.793,
      "step": 1367
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.4619673025408726,
      "learning_rate": 0.00032237709187839996,
      "loss": 2.9258,
      "step": 1368
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.48185254768727653,
      "learning_rate": 0.0003216830385021388,
      "loss": 3.0469,
      "step": 1369
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.4863470257084245,
      "learning_rate": 0.0003209893785595959,
      "loss": 2.9238,
      "step": 1370
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.4696792075065356,
      "learning_rate": 0.00032029611358124366,
      "loss": 2.9648,
      "step": 1371
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.4790003610668366,
      "learning_rate": 0.00031960324509668336,
      "loss": 2.8906,
      "step": 1372
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.4955695179175664,
      "learning_rate": 0.0003189107746346412,
      "loss": 2.9473,
      "step": 1373
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5280016629068969,
      "learning_rate": 0.0003182187037229653,
      "loss": 2.9062,
      "step": 1374
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5135127545092729,
      "learning_rate": 0.0003175270338886221,
      "loss": 2.9316,
      "step": 1375
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.4761023145418085,
      "learning_rate": 0.00031683576665769345,
      "loss": 2.8555,
      "step": 1376
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.4582033679053776,
      "learning_rate": 0.0003161449035553724,
      "loss": 2.916,
      "step": 1377
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.43713848387261833,
      "learning_rate": 0.00031545444610596077,
      "loss": 2.8691,
      "step": 1378
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5172508473020866,
      "learning_rate": 0.000314764395832865,
      "loss": 2.9336,
      "step": 1379
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5080767875693099,
      "learning_rate": 0.0003140747542585934,
      "loss": 2.8828,
      "step": 1380
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.4641824752538835,
      "learning_rate": 0.00031338552290475266,
      "loss": 2.8848,
      "step": 1381
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.48205806164544956,
      "learning_rate": 0.00031269670329204396,
      "loss": 2.8809,
      "step": 1382
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.47993581572997807,
      "learning_rate": 0.0003120082969402604,
      "loss": 2.9004,
      "step": 1383
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.503524808537732,
      "learning_rate": 0.00031132030536828314,
      "loss": 2.9824,
      "step": 1384
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4855241409925257,
      "learning_rate": 0.00031063273009407805,
      "loss": 2.9414,
      "step": 1385
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4770423210979097,
      "learning_rate": 0.00030994557263469265,
      "loss": 2.9297,
      "step": 1386
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4738218059159216,
      "learning_rate": 0.0003092588345062526,
      "loss": 2.9043,
      "step": 1387
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.46657017294587905,
      "learning_rate": 0.0003085725172239582,
      "loss": 2.9023,
      "step": 1388
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.48083190210032367,
      "learning_rate": 0.0003078866223020815,
      "loss": 2.9883,
      "step": 1389
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.48545936285814784,
      "learning_rate": 0.0003072011512539624,
      "loss": 2.8828,
      "step": 1390
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5417707347373938,
      "learning_rate": 0.00030651610559200574,
      "loss": 2.8535,
      "step": 1391
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.48456799257644484,
      "learning_rate": 0.00030583148682767757,
      "loss": 2.9004,
      "step": 1392
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4922116052310517,
      "learning_rate": 0.00030514729647150243,
      "loss": 2.9062,
      "step": 1393
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.47342899385267745,
      "learning_rate": 0.0003044635360330592,
      "loss": 2.875,
      "step": 1394
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4693354065495606,
      "learning_rate": 0.00030378020702097845,
      "loss": 2.8633,
      "step": 1395
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.45708458960574166,
      "learning_rate": 0.000303097310942939,
      "loss": 2.7988,
      "step": 1396
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5149364345362256,
      "learning_rate": 0.0003024148493056641,
      "loss": 2.9043,
      "step": 1397
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.470081840044759,
      "learning_rate": 0.00030173282361491865,
      "loss": 2.9512,
      "step": 1398
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4917907084288336,
      "learning_rate": 0.0003010512353755057,
      "loss": 2.9629,
      "step": 1399
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4962090960712131,
      "learning_rate": 0.00030037008609126313,
      "loss": 2.9355,
      "step": 1400
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.47856143624473113,
      "learning_rate": 0.0002996893772650602,
      "loss": 2.9395,
      "step": 1401
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.44096223423100794,
      "learning_rate": 0.0002990091103987945,
      "loss": 2.957,
      "step": 1402
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5766942159424121,
      "learning_rate": 0.0002983292869933886,
      "loss": 3.0176,
      "step": 1403
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4626443264931049,
      "learning_rate": 0.0002976499085487862,
      "loss": 2.9043,
      "step": 1404
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.47597646290373435,
      "learning_rate": 0.00029697097656394963,
      "loss": 2.8906,
      "step": 1405
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.4724973154174544,
      "learning_rate": 0.00029629249253685595,
      "loss": 2.8652,
      "step": 1406
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4723816935163741,
      "learning_rate": 0.00029561445796449416,
      "loss": 2.9141,
      "step": 1407
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4680408937184049,
      "learning_rate": 0.0002949368743428612,
      "loss": 2.7969,
      "step": 1408
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.5296869713249986,
      "learning_rate": 0.0002942597431669593,
      "loss": 2.9004,
      "step": 1409
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.48350028448613636,
      "learning_rate": 0.0002935830659307924,
      "loss": 2.9336,
      "step": 1410
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.5039051759601206,
      "learning_rate": 0.0002929068441273629,
      "loss": 2.8809,
      "step": 1411
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.5074470437195439,
      "learning_rate": 0.0002922310792486681,
      "loss": 2.875,
      "step": 1412
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.5020801286341401,
      "learning_rate": 0.00029155577278569745,
      "loss": 2.8418,
      "step": 1413
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4924181804297289,
      "learning_rate": 0.00029088092622842895,
      "loss": 2.8574,
      "step": 1414
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.479306182138484,
      "learning_rate": 0.00029020654106582544,
      "loss": 2.9336,
      "step": 1415
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.5323017302418239,
      "learning_rate": 0.0002895326187858326,
      "loss": 2.9062,
      "step": 1416
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.48288741651248906,
      "learning_rate": 0.00028885916087537377,
      "loss": 2.8281,
      "step": 1417
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4634639161873178,
      "learning_rate": 0.00028818616882034877,
      "loss": 2.9062,
      "step": 1418
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.5744880927297177,
      "learning_rate": 0.0002875136441056286,
      "loss": 2.8516,
      "step": 1419
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4925267538625905,
      "learning_rate": 0.000286841588215054,
      "loss": 2.8574,
      "step": 1420
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.45735540370366334,
      "learning_rate": 0.0002861700026314308,
      "loss": 2.9199,
      "step": 1421
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.49903120056550865,
      "learning_rate": 0.00028549888883652686,
      "loss": 2.9434,
      "step": 1422
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4594219937527614,
      "learning_rate": 0.00028482824831107,
      "loss": 2.8066,
      "step": 1423
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4933477046536055,
      "learning_rate": 0.000284158082534743,
      "loss": 2.9238,
      "step": 1424
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4713697353273177,
      "learning_rate": 0.00028348839298618177,
      "loss": 2.8711,
      "step": 1425
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4480252293413943,
      "learning_rate": 0.0002828191811429709,
      "loss": 2.8789,
      "step": 1426
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4622438405474986,
      "learning_rate": 0.00028215044848164164,
      "loss": 2.9375,
      "step": 1427
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.4894677459686669,
      "learning_rate": 0.00028148219647766747,
      "loss": 2.8984,
      "step": 1428
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5613692757205202,
      "learning_rate": 0.00028081442660546124,
      "loss": 2.9648,
      "step": 1429
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4750247327936129,
      "learning_rate": 0.0002801471403383728,
      "loss": 2.8555,
      "step": 1430
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.46658928490687523,
      "learning_rate": 0.00027948033914868415,
      "loss": 2.8281,
      "step": 1431
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.449160655751108,
      "learning_rate": 0.00027881402450760775,
      "loss": 2.8848,
      "step": 1432
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4780891840743034,
      "learning_rate": 0.00027814819788528165,
      "loss": 2.8691,
      "step": 1433
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4877829851979882,
      "learning_rate": 0.00027748286075076836,
      "loss": 2.9316,
      "step": 1434
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.49836271523384684,
      "learning_rate": 0.00027681801457204937,
      "loss": 2.8789,
      "step": 1435
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5325908709489241,
      "learning_rate": 0.00027615366081602306,
      "loss": 2.9141,
      "step": 1436
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5049062368702258,
      "learning_rate": 0.0002754898009485021,
      "loss": 2.9844,
      "step": 1437
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4849476477662056,
      "learning_rate": 0.0002748264364342085,
      "loss": 2.8613,
      "step": 1438
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4364775992881706,
      "learning_rate": 0.00027416356873677204,
      "loss": 2.9473,
      "step": 1439
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4722706451400139,
      "learning_rate": 0.0002735011993187258,
      "loss": 2.9004,
      "step": 1440
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5197905854998222,
      "learning_rate": 0.0002728393296415042,
      "loss": 2.8105,
      "step": 1441
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4771330216160533,
      "learning_rate": 0.00027217796116543817,
      "loss": 2.8594,
      "step": 1442
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5164037026526099,
      "learning_rate": 0.0002715170953497532,
      "loss": 2.8867,
      "step": 1443
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5062651206562234,
      "learning_rate": 0.00027085673365256614,
      "loss": 2.8555,
      "step": 1444
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5883003323371924,
      "learning_rate": 0.00027019687753088075,
      "loss": 2.8613,
      "step": 1445
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.48485775179221474,
      "learning_rate": 0.00026953752844058597,
      "loss": 2.9375,
      "step": 1446
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.47816948298325784,
      "learning_rate": 0.0002688786878364516,
      "loss": 2.8477,
      "step": 1447
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.477195438519884,
      "learning_rate": 0.00026822035717212597,
      "loss": 2.8125,
      "step": 1448
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4686372713609651,
      "learning_rate": 0.00026756253790013193,
      "loss": 2.9297,
      "step": 1449
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.5034271158119051,
      "learning_rate": 0.0002669052314718641,
      "loss": 2.8242,
      "step": 1450
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.4360252421761554,
      "learning_rate": 0.0002662484393375855,
      "loss": 2.9121,
      "step": 1451
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.45433283131802743,
      "learning_rate": 0.00026559216294642446,
      "loss": 2.9141,
      "step": 1452
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5034306225240862,
      "learning_rate": 0.0002649364037463718,
      "loss": 2.9473,
      "step": 1453
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.43947982389090195,
      "learning_rate": 0.0002642811631842764,
      "loss": 2.8984,
      "step": 1454
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.4647994914442885,
      "learning_rate": 0.0002636264427058439,
      "loss": 2.9102,
      "step": 1455
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5207193573900919,
      "learning_rate": 0.00026297224375563123,
      "loss": 2.9805,
      "step": 1456
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5022750450831048,
      "learning_rate": 0.00026231856777704575,
      "loss": 2.8652,
      "step": 1457
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.4522439370027197,
      "learning_rate": 0.00026166541621234026,
      "loss": 2.875,
      "step": 1458
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.42362366271998153,
      "learning_rate": 0.00026101279050261045,
      "loss": 2.9199,
      "step": 1459
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.4998811150543671,
      "learning_rate": 0.00026036069208779247,
      "loss": 2.9004,
      "step": 1460
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.492726639492851,
      "learning_rate": 0.0002597091224066581,
      "loss": 2.8418,
      "step": 1461
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.537572257087675,
      "learning_rate": 0.00025905808289681365,
      "loss": 2.8887,
      "step": 1462
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.49386000672368086,
      "learning_rate": 0.0002584075749946946,
      "loss": 2.8145,
      "step": 1463
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.47162634387976166,
      "learning_rate": 0.00025775760013556424,
      "loss": 2.9355,
      "step": 1464
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.52795817020811,
      "learning_rate": 0.0002571081597535095,
      "loss": 2.8281,
      "step": 1465
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5058284361939669,
      "learning_rate": 0.00025645925528143776,
      "loss": 2.8477,
      "step": 1466
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5346125956715768,
      "learning_rate": 0.0002558108881510747,
      "loss": 2.9766,
      "step": 1467
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5032674862134797,
      "learning_rate": 0.00025516305979295963,
      "loss": 2.9121,
      "step": 1468
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.45138364172770357,
      "learning_rate": 0.0002545157716364439,
      "loss": 2.8926,
      "step": 1469
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.504594007535425,
      "learning_rate": 0.00025386902510968624,
      "loss": 2.8672,
      "step": 1470
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.5268012321947474,
      "learning_rate": 0.00025322282163965095,
      "loss": 2.9219,
      "step": 1471
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.45852287754309506,
      "learning_rate": 0.00025257716265210384,
      "loss": 2.8809,
      "step": 1472
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.494739270648271,
      "learning_rate": 0.0002519320495716091,
      "loss": 2.8457,
      "step": 1473
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.457966348475691,
      "learning_rate": 0.00025128748382152716,
      "loss": 2.918,
      "step": 1474
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4705832497103963,
      "learning_rate": 0.00025064346682401016,
      "loss": 2.8145,
      "step": 1475
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.50081881558306,
      "learning_rate": 0.0002500000000000001,
      "loss": 2.9844,
      "step": 1476
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.47353509157237966,
      "learning_rate": 0.0002493570847692246,
      "loss": 2.8398,
      "step": 1477
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5420814076116663,
      "learning_rate": 0.00024871472255019424,
      "loss": 2.8867,
      "step": 1478
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5417868671523444,
      "learning_rate": 0.00024807291476019994,
      "loss": 2.9492,
      "step": 1479
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5162737216026438,
      "learning_rate": 0.00024743166281530877,
      "loss": 2.9004,
      "step": 1480
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4801799371158688,
      "learning_rate": 0.000246790968130362,
      "loss": 2.9121,
      "step": 1481
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4726305287875745,
      "learning_rate": 0.0002461508321189706,
      "loss": 2.9023,
      "step": 1482
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.47308354693956217,
      "learning_rate": 0.00024551125619351385,
      "loss": 2.875,
      "step": 1483
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.46134863680367977,
      "learning_rate": 0.00024487224176513453,
      "loss": 2.959,
      "step": 1484
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4952192076793737,
      "learning_rate": 0.0002442337902437365,
      "loss": 2.8555,
      "step": 1485
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.46338704783379,
      "learning_rate": 0.0002435959030379824,
      "loss": 2.8926,
      "step": 1486
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4837456911419315,
      "learning_rate": 0.00024295858155528888,
      "loss": 2.9199,
      "step": 1487
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5102861235112165,
      "learning_rate": 0.00024232182720182523,
      "loss": 2.8906,
      "step": 1488
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.521544170043346,
      "learning_rate": 0.00024168564138250855,
      "loss": 2.9434,
      "step": 1489
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4984521562069893,
      "learning_rate": 0.00024105002550100246,
      "loss": 2.9277,
      "step": 1490
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4905257038565216,
      "learning_rate": 0.00024041498095971254,
      "loss": 2.8594,
      "step": 1491
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4767754120178332,
      "learning_rate": 0.0002397805091597835,
      "loss": 2.9082,
      "step": 1492
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5216412792072983,
      "learning_rate": 0.0002391466115010973,
      "loss": 2.9238,
      "step": 1493
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.578231949786693,
      "learning_rate": 0.00023851328938226808,
      "loss": 2.8359,
      "step": 1494
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.49574749956781283,
      "learning_rate": 0.00023788054420064109,
      "loss": 2.9316,
      "step": 1495
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.46179559810474924,
      "learning_rate": 0.00023724837735228773,
      "loss": 2.9004,
      "step": 1496
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.4833458521145663,
      "learning_rate": 0.00023661679023200422,
      "loss": 2.9082,
      "step": 1497
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.4837850460615172,
      "learning_rate": 0.00023598578423330714,
      "loss": 2.8496,
      "step": 1498
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.47001298485091975,
      "learning_rate": 0.00023535536074843083,
      "loss": 2.9336,
      "step": 1499
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.46514713077151265,
      "learning_rate": 0.00023472552116832502,
      "loss": 2.873,
      "step": 1500
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.4770948807920571,
      "learning_rate": 0.0002340962668826503,
      "loss": 2.8086,
      "step": 1501
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.46689429959844164,
      "learning_rate": 0.00023346759927977663,
      "loss": 2.8477,
      "step": 1502
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.4694353480563685,
      "learning_rate": 0.0002328395197467789,
      "loss": 2.8574,
      "step": 1503
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.48017292870045086,
      "learning_rate": 0.00023221202966943515,
      "loss": 2.9434,
      "step": 1504
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5204708018438906,
      "learning_rate": 0.0002315851304322223,
      "loss": 2.9727,
      "step": 1505
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5490070797698121,
      "learning_rate": 0.0002309588234183137,
      "loss": 2.8711,
      "step": 1506
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.45728022303092514,
      "learning_rate": 0.00023033311000957653,
      "loss": 2.8594,
      "step": 1507
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5263416003679033,
      "learning_rate": 0.00022970799158656758,
      "loss": 2.8672,
      "step": 1508
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.4644926462210496,
      "learning_rate": 0.0002290834695285316,
      "loss": 2.9355,
      "step": 1509
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5182895961629406,
      "learning_rate": 0.00022845954521339678,
      "loss": 2.8574,
      "step": 1510
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5037923216132947,
      "learning_rate": 0.0002278362200177732,
      "loss": 2.8848,
      "step": 1511
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.47343420809459597,
      "learning_rate": 0.00022721349531694852,
      "loss": 2.8281,
      "step": 1512
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.5151539945982408,
      "learning_rate": 0.0002265913724848855,
      "loss": 2.9023,
      "step": 1513
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.4711313689048708,
      "learning_rate": 0.00022596985289421946,
      "loss": 2.9766,
      "step": 1514
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.46371666470664374,
      "learning_rate": 0.00022534893791625405,
      "loss": 2.8125,
      "step": 1515
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5067884210682959,
      "learning_rate": 0.00022472862892095968,
      "loss": 2.7773,
      "step": 1516
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5005133987439213,
      "learning_rate": 0.00022410892727696896,
      "loss": 2.9219,
      "step": 1517
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5219737039302164,
      "learning_rate": 0.0002234898343515751,
      "loss": 2.916,
      "step": 1518
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4372954097110122,
      "learning_rate": 0.00022287135151072792,
      "loss": 2.7754,
      "step": 1519
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.46420311368224565,
      "learning_rate": 0.00022225348011903096,
      "loss": 2.8496,
      "step": 1520
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4614535422400789,
      "learning_rate": 0.0002216362215397393,
      "loss": 2.8203,
      "step": 1521
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.48448135065021547,
      "learning_rate": 0.00022101957713475522,
      "loss": 2.9922,
      "step": 1522
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4642449320986145,
      "learning_rate": 0.00022040354826462666,
      "loss": 2.9414,
      "step": 1523
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4620421382760685,
      "learning_rate": 0.0002197881362885426,
      "loss": 2.8105,
      "step": 1524
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4899016377072257,
      "learning_rate": 0.0002191733425643318,
      "loss": 2.8555,
      "step": 1525
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4835555432699328,
      "learning_rate": 0.00021855916844845826,
      "loss": 2.8906,
      "step": 1526
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4879024044440522,
      "learning_rate": 0.00021794561529601898,
      "loss": 2.9102,
      "step": 1527
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4319093428622175,
      "learning_rate": 0.00021733268446074138,
      "loss": 2.916,
      "step": 1528
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4992094435982433,
      "learning_rate": 0.00021672037729497917,
      "loss": 2.9395,
      "step": 1529
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.48018282745890606,
      "learning_rate": 0.0002161086951497106,
      "loss": 2.8535,
      "step": 1530
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5386443091974059,
      "learning_rate": 0.00021549763937453442,
      "loss": 2.8281,
      "step": 1531
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.4967447263333056,
      "learning_rate": 0.00021488721131766736,
      "loss": 2.8574,
      "step": 1532
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5125048497621292,
      "learning_rate": 0.00021427741232594183,
      "loss": 2.9199,
      "step": 1533
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.517946110616049,
      "learning_rate": 0.0002136682437448013,
      "loss": 2.9473,
      "step": 1534
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.49172418790452305,
      "learning_rate": 0.0002130597069182994,
      "loss": 2.8184,
      "step": 1535
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5110416340560899,
      "learning_rate": 0.0002124518031890948,
      "loss": 2.8359,
      "step": 1536
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.49435926489348175,
      "learning_rate": 0.0002118445338984502,
      "loss": 2.9082,
      "step": 1537
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.5049020622716153,
      "learning_rate": 0.00021123790038622808,
      "loss": 2.7988,
      "step": 1538
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.5480671038844431,
      "learning_rate": 0.0002106319039908879,
      "loss": 2.9141,
      "step": 1539
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.510065226200887,
      "learning_rate": 0.00021002654604948412,
      "loss": 2.9297,
      "step": 1540
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.46967583469014024,
      "learning_rate": 0.00020942182789766172,
      "loss": 2.8516,
      "step": 1541
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.45214833528381576,
      "learning_rate": 0.00020881775086965492,
      "loss": 2.832,
      "step": 1542
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.4825676278445836,
      "learning_rate": 0.00020821431629828246,
      "loss": 2.8262,
      "step": 1543
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.49639521880262527,
      "learning_rate": 0.00020761152551494643,
      "loss": 3.0254,
      "step": 1544
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.46672708234506455,
      "learning_rate": 0.00020700937984962798,
      "loss": 2.8398,
      "step": 1545
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.45782204745330934,
      "learning_rate": 0.0002064078806308848,
      "loss": 2.8379,
      "step": 1546
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.49024752981794767,
      "learning_rate": 0.00020580702918584882,
      "loss": 2.9277,
      "step": 1547
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.4399875104746515,
      "learning_rate": 0.000205206826840222,
      "loss": 2.8848,
      "step": 1548
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.46446407435822706,
      "learning_rate": 0.0002046072749182751,
      "loss": 2.8477,
      "step": 1549
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.4932448237807614,
      "learning_rate": 0.00020400837474284273,
      "loss": 2.8047,
      "step": 1550
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.4596241002492523,
      "learning_rate": 0.0002034101276353224,
      "loss": 2.8438,
      "step": 1551
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.5080460631393607,
      "learning_rate": 0.00020281253491567027,
      "loss": 2.8887,
      "step": 1552
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.5694892462291503,
      "learning_rate": 0.0002022155979023984,
      "loss": 2.9688,
      "step": 1553
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.47454837001788175,
      "learning_rate": 0.000201619317912573,
      "loss": 2.8984,
      "step": 1554
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.5224391376769193,
      "learning_rate": 0.00020102369626180962,
      "loss": 2.9297,
      "step": 1555
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.4524193409768758,
      "learning_rate": 0.0002004287342642721,
      "loss": 2.8867,
      "step": 1556
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.5074174101129643,
      "learning_rate": 0.00019983443323266824,
      "loss": 2.8242,
      "step": 1557
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.46980280530638685,
      "learning_rate": 0.00019924079447824805,
      "loss": 2.8652,
      "step": 1558
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.49542597461783294,
      "learning_rate": 0.00019864781931079977,
      "loss": 2.918,
      "step": 1559
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.45967276022807413,
      "learning_rate": 0.00019805550903864773,
      "loss": 2.9531,
      "step": 1560
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.4968576604063848,
      "learning_rate": 0.00019746386496864948,
      "loss": 2.9199,
      "step": 1561
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.43423566728299606,
      "learning_rate": 0.00019687288840619226,
      "loss": 2.8945,
      "step": 1562
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5949370047324524,
      "learning_rate": 0.0001962825806551911,
      "loss": 2.9199,
      "step": 1563
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.48343666616671055,
      "learning_rate": 0.0001956929430180846,
      "loss": 2.8965,
      "step": 1564
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.46035672179173637,
      "learning_rate": 0.00019510397679583374,
      "loss": 2.8281,
      "step": 1565
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.49496526581091227,
      "learning_rate": 0.0001945156832879174,
      "loss": 2.8242,
      "step": 1566
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5070666177425076,
      "learning_rate": 0.00019392806379233036,
      "loss": 2.9844,
      "step": 1567
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5078238115174167,
      "learning_rate": 0.00019334111960558065,
      "loss": 2.8809,
      "step": 1568
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.4626300645609412,
      "learning_rate": 0.00019275485202268573,
      "loss": 2.8906,
      "step": 1569
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.505288211042878,
      "learning_rate": 0.00019216926233717085,
      "loss": 2.8281,
      "step": 1570
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.49747245492055,
      "learning_rate": 0.00019158435184106498,
      "loss": 2.9277,
      "step": 1571
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5314619220922183,
      "learning_rate": 0.00019100012182489905,
      "loss": 2.8555,
      "step": 1572
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5944358737941624,
      "learning_rate": 0.00019041657357770226,
      "loss": 2.8184,
      "step": 1573
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.4949249029631447,
      "learning_rate": 0.00018983370838699943,
      "loss": 2.8984,
      "step": 1574
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5327522585279723,
      "learning_rate": 0.00018925152753880892,
      "loss": 2.9629,
      "step": 1575
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5112629575719262,
      "learning_rate": 0.00018867003231763847,
      "loss": 2.8555,
      "step": 1576
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5496561898408718,
      "learning_rate": 0.00018808922400648375,
      "loss": 2.9531,
      "step": 1577
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.4864102787849932,
      "learning_rate": 0.00018750910388682428,
      "loss": 2.9336,
      "step": 1578
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.46611130570337295,
      "learning_rate": 0.00018692967323862125,
      "loss": 2.832,
      "step": 1579
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.4748327608647628,
      "learning_rate": 0.00018635093334031517,
      "loss": 2.8281,
      "step": 1580
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.49569442753981185,
      "learning_rate": 0.00018577288546882165,
      "loss": 2.8789,
      "step": 1581
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.49035154586634855,
      "learning_rate": 0.00018519553089953023,
      "loss": 2.9004,
      "step": 1582
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.5340750119237502,
      "learning_rate": 0.0001846188709063001,
      "loss": 2.9434,
      "step": 1583
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4770331410909047,
      "learning_rate": 0.00018404290676145857,
      "loss": 2.8027,
      "step": 1584
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.47475013582088293,
      "learning_rate": 0.00018346763973579722,
      "loss": 2.8945,
      "step": 1585
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.45849814807438727,
      "learning_rate": 0.00018289307109856939,
      "loss": 2.9609,
      "step": 1586
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.5379858844519845,
      "learning_rate": 0.0001823192021174882,
      "loss": 2.8359,
      "step": 1587
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.473879398290997,
      "learning_rate": 0.0001817460340587223,
      "loss": 2.8613,
      "step": 1588
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4771731862876735,
      "learning_rate": 0.00018117356818689445,
      "loss": 2.8828,
      "step": 1589
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4500931370248715,
      "learning_rate": 0.00018060180576507756,
      "loss": 2.8105,
      "step": 1590
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4316439141659538,
      "learning_rate": 0.00018003074805479313,
      "loss": 2.791,
      "step": 1591
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4810214391052127,
      "learning_rate": 0.00017946039631600724,
      "loss": 2.9102,
      "step": 1592
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.541483451473701,
      "learning_rate": 0.00017889075180712837,
      "loss": 2.9395,
      "step": 1593
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.5315088047533172,
      "learning_rate": 0.00017832181578500512,
      "loss": 2.7305,
      "step": 1594
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4610097831486203,
      "learning_rate": 0.0001777535895049221,
      "loss": 2.9688,
      "step": 1595
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.46007863285189016,
      "learning_rate": 0.0001771860742205988,
      "loss": 2.9141,
      "step": 1596
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.5134015803924606,
      "learning_rate": 0.00017661927118418525,
      "loss": 2.8965,
      "step": 1597
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4870923373416817,
      "learning_rate": 0.00017605318164626066,
      "loss": 2.8535,
      "step": 1598
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.5041719189589399,
      "learning_rate": 0.00017548780685582949,
      "loss": 2.8984,
      "step": 1599
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4998305744288629,
      "learning_rate": 0.00017492314806031922,
      "loss": 2.916,
      "step": 1600
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4824223241762916,
      "learning_rate": 0.00017435920650557806,
      "loss": 2.8359,
      "step": 1601
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.45777903642300666,
      "learning_rate": 0.00017379598343587112,
      "loss": 2.8398,
      "step": 1602
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.46253182023819717,
      "learning_rate": 0.00017323348009387878,
      "loss": 2.8477,
      "step": 1603
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.46753582296224006,
      "learning_rate": 0.0001726716977206929,
      "loss": 2.8926,
      "step": 1604
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.515523525807142,
      "learning_rate": 0.00017211063755581525,
      "loss": 2.9668,
      "step": 1605
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.49711537777497894,
      "learning_rate": 0.0001715503008371536,
      "loss": 2.8516,
      "step": 1606
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5539428673655535,
      "learning_rate": 0.0001709906888010196,
      "loss": 2.9766,
      "step": 1607
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.48399754059577893,
      "learning_rate": 0.00017043180268212638,
      "loss": 2.8984,
      "step": 1608
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5051378870739877,
      "learning_rate": 0.00016987364371358481,
      "loss": 2.8516,
      "step": 1609
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5363845034095245,
      "learning_rate": 0.00016931621312690214,
      "loss": 3.0098,
      "step": 1610
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5201292186445707,
      "learning_rate": 0.00016875951215197777,
      "loss": 2.8281,
      "step": 1611
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4783932063370838,
      "learning_rate": 0.00016820354201710214,
      "loss": 2.959,
      "step": 1612
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4489302947809322,
      "learning_rate": 0.00016764830394895203,
      "loss": 2.8379,
      "step": 1613
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4456015985812711,
      "learning_rate": 0.00016709379917259027,
      "loss": 2.8633,
      "step": 1614
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4920082942759353,
      "learning_rate": 0.00016654002891146091,
      "loss": 2.8652,
      "step": 1615
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5018120368062529,
      "learning_rate": 0.00016598699438738764,
      "loss": 2.8945,
      "step": 1616
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5207737954638507,
      "learning_rate": 0.00016543469682057105,
      "loss": 2.9531,
      "step": 1617
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.47615305889654813,
      "learning_rate": 0.00016488313742958526,
      "loss": 2.8379,
      "step": 1618
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5284723700511739,
      "learning_rate": 0.00016433231743137646,
      "loss": 2.8945,
      "step": 1619
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4947507051723159,
      "learning_rate": 0.0001637822380412584,
      "loss": 2.8672,
      "step": 1620
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5042855374873858,
      "learning_rate": 0.00016323290047291195,
      "loss": 2.9258,
      "step": 1621
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4869086834595209,
      "learning_rate": 0.0001626843059383803,
      "loss": 2.9141,
      "step": 1622
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5051106724689701,
      "learning_rate": 0.00016213645564806752,
      "loss": 2.8867,
      "step": 1623
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.48225293710805744,
      "learning_rate": 0.0001615893508107359,
      "loss": 2.9238,
      "step": 1624
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4468080775511376,
      "learning_rate": 0.00016104299263350252,
      "loss": 2.8164,
      "step": 1625
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.48398787575357954,
      "learning_rate": 0.00016049738232183758,
      "loss": 2.8887,
      "step": 1626
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.48248791345201436,
      "learning_rate": 0.0001599525210795606,
      "loss": 2.8184,
      "step": 1627
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5188149016078198,
      "learning_rate": 0.00015940841010883889,
      "loss": 2.877,
      "step": 1628
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5145246417185223,
      "learning_rate": 0.00015886505061018413,
      "loss": 2.9316,
      "step": 1629
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4995091784087288,
      "learning_rate": 0.0001583224437824498,
      "loss": 2.9004,
      "step": 1630
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.47109545814196974,
      "learning_rate": 0.0001577805908228293,
      "loss": 2.8477,
      "step": 1631
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4621542681311392,
      "learning_rate": 0.00015723949292685191,
      "loss": 2.8789,
      "step": 1632
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.49296207829880645,
      "learning_rate": 0.0001566991512883818,
      "loss": 2.8418,
      "step": 1633
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.49050082665810407,
      "learning_rate": 0.00015615956709961378,
      "loss": 2.873,
      "step": 1634
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5301554610212675,
      "learning_rate": 0.00015562074155107215,
      "loss": 2.9668,
      "step": 1635
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4935109521318676,
      "learning_rate": 0.0001550826758316068,
      "loss": 2.9316,
      "step": 1636
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4753622148080949,
      "learning_rate": 0.00015454537112839122,
      "loss": 2.9375,
      "step": 1637
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4933351520333699,
      "learning_rate": 0.00015400882862692033,
      "loss": 2.8184,
      "step": 1638
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4695663054235339,
      "learning_rate": 0.00015347304951100665,
      "loss": 2.832,
      "step": 1639
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.48723869071526266,
      "learning_rate": 0.00015293803496277907,
      "loss": 2.8418,
      "step": 1640
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.48440460626043846,
      "learning_rate": 0.00015240378616267886,
      "loss": 2.9238,
      "step": 1641
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5015692103727495,
      "learning_rate": 0.00015187030428945843,
      "loss": 2.9141,
      "step": 1642
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.47705366553871503,
      "learning_rate": 0.0001513375905201776,
      "loss": 2.8242,
      "step": 1643
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5013748399721134,
      "learning_rate": 0.00015080564603020142,
      "loss": 2.9629,
      "step": 1644
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.46892131326727277,
      "learning_rate": 0.0001502744719931982,
      "loss": 2.8672,
      "step": 1645
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4998302615147639,
      "learning_rate": 0.00014974406958113558,
      "loss": 2.9766,
      "step": 1646
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5149775902226792,
      "learning_rate": 0.00014921443996427947,
      "loss": 2.9316,
      "step": 1647
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5164167890283942,
      "learning_rate": 0.0001486855843111901,
      "loss": 2.8203,
      "step": 1648
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.46704163360604667,
      "learning_rate": 0.0001481575037887201,
      "loss": 2.9512,
      "step": 1649
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.523970802653406,
      "learning_rate": 0.00014763019956201253,
      "loss": 2.8477,
      "step": 1650
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5130817084093677,
      "learning_rate": 0.0001471036727944966,
      "loss": 2.9023,
      "step": 1651
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.49226610748914673,
      "learning_rate": 0.0001465779246478872,
      "loss": 2.8223,
      "step": 1652
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.48904423277760994,
      "learning_rate": 0.00014605295628218045,
      "loss": 2.9688,
      "step": 1653
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.48768902197093383,
      "learning_rate": 0.0001455287688556527,
      "loss": 2.8613,
      "step": 1654
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.46338867785428123,
      "learning_rate": 0.00014500536352485673,
      "loss": 2.8652,
      "step": 1655
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.4927611904336579,
      "learning_rate": 0.00014448274144461965,
      "loss": 2.9707,
      "step": 1656
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5349720213378326,
      "learning_rate": 0.00014396090376804112,
      "loss": 2.9375,
      "step": 1657
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.4585149970781446,
      "learning_rate": 0.00014343985164648926,
      "loss": 2.7832,
      "step": 1658
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5113757246156228,
      "learning_rate": 0.00014291958622959973,
      "loss": 2.832,
      "step": 1659
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.505262017111781,
      "learning_rate": 0.00014240010866527176,
      "loss": 2.8496,
      "step": 1660
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.47581856869282596,
      "learning_rate": 0.00014188142009966686,
      "loss": 2.8516,
      "step": 1661
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.49160906849326663,
      "learning_rate": 0.0001413635216772053,
      "loss": 2.8184,
      "step": 1662
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.4688256902357146,
      "learning_rate": 0.000140846414540564,
      "loss": 2.7832,
      "step": 1663
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.45239317159773323,
      "learning_rate": 0.00014033009983067452,
      "loss": 2.9219,
      "step": 1664
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5078089493274349,
      "learning_rate": 0.00013981457868671927,
      "loss": 2.9238,
      "step": 1665
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.4668382347073814,
      "learning_rate": 0.0001392998522461305,
      "loss": 2.8848,
      "step": 1666
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5314633559194283,
      "learning_rate": 0.00013878592164458635,
      "loss": 2.8828,
      "step": 1667
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.47856704428958646,
      "learning_rate": 0.00013827278801600978,
      "loss": 2.8242,
      "step": 1668
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.4868051107996366,
      "learning_rate": 0.0001377604524925647,
      "loss": 2.9434,
      "step": 1669
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.4821601355334815,
      "learning_rate": 0.00013724891620465424,
      "loss": 2.8516,
      "step": 1670
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5356696882204088,
      "learning_rate": 0.0001367381802809185,
      "loss": 2.7969,
      "step": 1671
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.48930281059909153,
      "learning_rate": 0.00013622824584823113,
      "loss": 2.8398,
      "step": 1672
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5200598325022292,
      "learning_rate": 0.00013571911403169795,
      "loss": 2.877,
      "step": 1673
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.45613615674132274,
      "learning_rate": 0.0001352107859546533,
      "loss": 2.8164,
      "step": 1674
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5028038306358791,
      "learning_rate": 0.00013470326273865886,
      "loss": 2.9141,
      "step": 1675
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5306314858154745,
      "learning_rate": 0.00013419654550349985,
      "loss": 2.8887,
      "step": 1676
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.468410151872843,
      "learning_rate": 0.00013369063536718346,
      "loss": 2.8438,
      "step": 1677
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.4969784002856183,
      "learning_rate": 0.00013318553344593632,
      "loss": 2.8203,
      "step": 1678
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5569122095821211,
      "learning_rate": 0.00013268124085420136,
      "loss": 2.9297,
      "step": 1679
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5575136386153504,
      "learning_rate": 0.0001321777587046364,
      "loss": 2.8672,
      "step": 1680
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5256119467109397,
      "learning_rate": 0.00013167508810811059,
      "loss": 2.873,
      "step": 1681
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.4580666326612865,
      "learning_rate": 0.0001311732301737029,
      "loss": 2.834,
      "step": 1682
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.49191232233134247,
      "learning_rate": 0.0001306721860086991,
      "loss": 2.877,
      "step": 1683
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5163619530481239,
      "learning_rate": 0.00013017195671858928,
      "loss": 2.873,
      "step": 1684
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5149807619903011,
      "learning_rate": 0.0001296725434070661,
      "loss": 2.9043,
      "step": 1685
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.499401504128304,
      "learning_rate": 0.00012917394717602121,
      "loss": 2.8711,
      "step": 1686
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.499183324062122,
      "learning_rate": 0.00012867616912554426,
      "loss": 2.9434,
      "step": 1687
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.49016248902128823,
      "learning_rate": 0.00012817921035391882,
      "loss": 2.875,
      "step": 1688
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5173182959639606,
      "learning_rate": 0.00012768307195762168,
      "loss": 2.9629,
      "step": 1689
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.45537803119207315,
      "learning_rate": 0.00012718775503131908,
      "loss": 2.7852,
      "step": 1690
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4862756421335543,
      "learning_rate": 0.0001266932606678646,
      "loss": 2.8672,
      "step": 1691
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4825853253558514,
      "learning_rate": 0.00012619958995829756,
      "loss": 2.875,
      "step": 1692
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.45663716296274753,
      "learning_rate": 0.0001257067439918394,
      "loss": 2.8047,
      "step": 1693
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4867159610447571,
      "learning_rate": 0.00012521472385589234,
      "loss": 2.8223,
      "step": 1694
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5223756658741734,
      "learning_rate": 0.00012472353063603626,
      "loss": 2.7695,
      "step": 1695
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5251198430000581,
      "learning_rate": 0.0001242331654160263,
      "loss": 2.9238,
      "step": 1696
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5078866374076179,
      "learning_rate": 0.0001237436292777914,
      "loss": 2.9531,
      "step": 1697
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4974425697836808,
      "learning_rate": 0.00012325492330143061,
      "loss": 2.8965,
      "step": 1698
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.49953102975724667,
      "learning_rate": 0.00012276704856521175,
      "loss": 2.8516,
      "step": 1699
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5197382694640409,
      "learning_rate": 0.00012228000614556816,
      "loss": 2.9824,
      "step": 1700
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5311437020670159,
      "learning_rate": 0.00012179379711709738,
      "loss": 2.9492,
      "step": 1701
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4926408673158625,
      "learning_rate": 0.0001213084225525577,
      "loss": 2.8516,
      "step": 1702
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.464877524195786,
      "learning_rate": 0.00012082388352286627,
      "loss": 2.8926,
      "step": 1703
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4650917634715133,
      "learning_rate": 0.00012034018109709716,
      "loss": 2.8477,
      "step": 1704
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.47929026222952487,
      "learning_rate": 0.00011985731634247809,
      "loss": 2.8379,
      "step": 1705
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.515407792629518,
      "learning_rate": 0.00011937529032438904,
      "loss": 2.9512,
      "step": 1706
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4772295152137335,
      "learning_rate": 0.00011889410410635887,
      "loss": 2.832,
      "step": 1707
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5523897721084357,
      "learning_rate": 0.0001184137587500641,
      "loss": 2.8945,
      "step": 1708
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.5032668072339842,
      "learning_rate": 0.00011793425531532564,
      "loss": 2.791,
      "step": 1709
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.47975691734180387,
      "learning_rate": 0.00011745559486010671,
      "loss": 2.8965,
      "step": 1710
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4788453938542716,
      "learning_rate": 0.00011697777844051105,
      "loss": 2.9141,
      "step": 1711
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.4928613899450833,
      "learning_rate": 0.00011650080711077964,
      "loss": 2.9805,
      "step": 1712
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.49623231792943395,
      "learning_rate": 0.00011602468192328936,
      "loss": 2.7832,
      "step": 1713
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5039287041346067,
      "learning_rate": 0.00011554940392854973,
      "loss": 2.8438,
      "step": 1714
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4852390761898434,
      "learning_rate": 0.00011507497417520146,
      "loss": 2.9004,
      "step": 1715
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4796375611606061,
      "learning_rate": 0.00011460139371001339,
      "loss": 2.8164,
      "step": 1716
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4932666927602591,
      "learning_rate": 0.00011412866357788049,
      "loss": 2.8457,
      "step": 1717
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.49225283725511687,
      "learning_rate": 0.00011365678482182207,
      "loss": 2.8594,
      "step": 1718
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4673651175099916,
      "learning_rate": 0.0001131857584829783,
      "loss": 2.8965,
      "step": 1719
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4596467343637949,
      "learning_rate": 0.0001127155856006093,
      "loss": 2.8477,
      "step": 1720
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.46508052202454997,
      "learning_rate": 0.00011224626721209141,
      "loss": 2.8984,
      "step": 1721
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4969708197852194,
      "learning_rate": 0.0001117778043529164,
      "loss": 2.9434,
      "step": 1722
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.488908720073959,
      "learning_rate": 0.0001113101980566879,
      "loss": 2.8398,
      "step": 1723
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4873984474859213,
      "learning_rate": 0.00011084344935511958,
      "loss": 2.9668,
      "step": 1724
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5104723138380299,
      "learning_rate": 0.00011037755927803345,
      "loss": 2.7363,
      "step": 1725
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4979790018385743,
      "learning_rate": 0.00010991252885335651,
      "loss": 2.916,
      "step": 1726
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5021356596201003,
      "learning_rate": 0.00010944835910711958,
      "loss": 2.9082,
      "step": 1727
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4757879039055125,
      "learning_rate": 0.00010898505106345396,
      "loss": 2.7109,
      "step": 1728
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5028959544799132,
      "learning_rate": 0.00010852260574459022,
      "loss": 2.9062,
      "step": 1729
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5019206512173225,
      "learning_rate": 0.00010806102417085512,
      "loss": 2.7695,
      "step": 1730
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4898397996292006,
      "learning_rate": 0.00010760030736066951,
      "loss": 2.8359,
      "step": 1731
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5414410731281782,
      "learning_rate": 0.00010714045633054687,
      "loss": 2.8594,
      "step": 1732
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5047891309589607,
      "learning_rate": 0.00010668147209508971,
      "loss": 2.8945,
      "step": 1733
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5249744725017079,
      "learning_rate": 0.00010622335566698877,
      "loss": 2.9023,
      "step": 1734
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5030488325625009,
      "learning_rate": 0.00010576610805701942,
      "loss": 2.8926,
      "step": 1735
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.49831673648265357,
      "learning_rate": 0.00010530973027404073,
      "loss": 2.8848,
      "step": 1736
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.48102954146056465,
      "learning_rate": 0.00010485422332499212,
      "loss": 2.8613,
      "step": 1737
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5060142235843171,
      "learning_rate": 0.00010439958821489165,
      "loss": 2.8105,
      "step": 1738
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5125470804322958,
      "learning_rate": 0.00010394582594683428,
      "loss": 2.873,
      "step": 1739
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5270344449639704,
      "learning_rate": 0.0001034929375219884,
      "loss": 2.8379,
      "step": 1740
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5140552160637826,
      "learning_rate": 0.00010304092393959514,
      "loss": 2.8926,
      "step": 1741
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.4870938247292708,
      "learning_rate": 0.00010258978619696468,
      "loss": 2.8379,
      "step": 1742
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.48816180633893286,
      "learning_rate": 0.00010213952528947551,
      "loss": 2.8496,
      "step": 1743
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.48576624959270626,
      "learning_rate": 0.00010169014221057089,
      "loss": 2.9375,
      "step": 1744
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5914185615261834,
      "learning_rate": 0.00010124163795175734,
      "loss": 2.7207,
      "step": 1745
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.4609337265180438,
      "learning_rate": 0.00010079401350260287,
      "loss": 2.8906,
      "step": 1746
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5136387194128935,
      "learning_rate": 0.00010034726985073362,
      "loss": 2.959,
      "step": 1747
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.4937350179907994,
      "learning_rate": 9.9901407981833e-05,
      "loss": 2.9648,
      "step": 1748
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.510104136298016,
      "learning_rate": 9.94564288796384e-05,
      "loss": 2.793,
      "step": 1749
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5043307748599032,
      "learning_rate": 9.901233352593953e-05,
      "loss": 2.9023,
      "step": 1750
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.49795132028914674,
      "learning_rate": 9.856912290057668e-05,
      "loss": 2.8184,
      "step": 1751
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.4958483221947375,
      "learning_rate": 9.812679798143748e-05,
      "loss": 2.7617,
      "step": 1752
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.49082809837857794,
      "learning_rate": 9.768535974445586e-05,
      "loss": 2.8301,
      "step": 1753
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5318990686848214,
      "learning_rate": 9.724480916360906e-05,
      "loss": 2.9121,
      "step": 1754
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.4858272635026834,
      "learning_rate": 9.68051472109162e-05,
      "loss": 2.8223,
      "step": 1755
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4935556810803597,
      "learning_rate": 9.636637485643529e-05,
      "loss": 2.8926,
      "step": 1756
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5465154797061601,
      "learning_rate": 9.592849306826174e-05,
      "loss": 2.9102,
      "step": 1757
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.49718700299957086,
      "learning_rate": 9.549150281252633e-05,
      "loss": 2.9316,
      "step": 1758
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5086339625036027,
      "learning_rate": 9.505540505339223e-05,
      "loss": 2.7949,
      "step": 1759
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.49865066376339745,
      "learning_rate": 9.4620200753054e-05,
      "loss": 2.8516,
      "step": 1760
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5097049575067071,
      "learning_rate": 9.418589087173441e-05,
      "loss": 2.8809,
      "step": 1761
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.49918589669966756,
      "learning_rate": 9.375247636768325e-05,
      "loss": 2.8848,
      "step": 1762
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5077042589519181,
      "learning_rate": 9.331995819717443e-05,
      "loss": 2.9414,
      "step": 1763
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5294372816760746,
      "learning_rate": 9.288833731450419e-05,
      "loss": 2.9473,
      "step": 1764
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4792584033054512,
      "learning_rate": 9.245761467198948e-05,
      "loss": 2.9062,
      "step": 1765
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5477950932729524,
      "learning_rate": 9.20277912199648e-05,
      "loss": 2.7949,
      "step": 1766
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4791766176067013,
      "learning_rate": 9.159886790678123e-05,
      "loss": 2.8184,
      "step": 1767
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5246794602790502,
      "learning_rate": 9.11708456788033e-05,
      "loss": 2.8574,
      "step": 1768
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.44707413815268277,
      "learning_rate": 9.074372548040793e-05,
      "loss": 2.959,
      "step": 1769
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.47069389456759475,
      "learning_rate": 9.031750825398145e-05,
      "loss": 2.8926,
      "step": 1770
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4945702275864182,
      "learning_rate": 8.98921949399179e-05,
      "loss": 2.8789,
      "step": 1771
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4623418090049629,
      "learning_rate": 8.94677864766173e-05,
      "loss": 2.9277,
      "step": 1772
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4988803561296559,
      "learning_rate": 8.904428380048269e-05,
      "loss": 2.8516,
      "step": 1773
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.48719876832983955,
      "learning_rate": 8.862168784591929e-05,
      "loss": 2.8027,
      "step": 1774
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4723560692140685,
      "learning_rate": 8.819999954533115e-05,
      "loss": 2.8613,
      "step": 1775
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.5054574011044716,
      "learning_rate": 8.777921982911996e-05,
      "loss": 2.9141,
      "step": 1776
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.48160829385398607,
      "learning_rate": 8.735934962568253e-05,
      "loss": 2.9121,
      "step": 1777
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4969687346354118,
      "learning_rate": 8.694038986140945e-05,
      "loss": 2.9707,
      "step": 1778
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.49677285504651936,
      "learning_rate": 8.652234146068206e-05,
      "loss": 2.832,
      "step": 1779
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5178288387306751,
      "learning_rate": 8.610520534587086e-05,
      "loss": 2.791,
      "step": 1780
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5247387485286996,
      "learning_rate": 8.568898243733397e-05,
      "loss": 2.8047,
      "step": 1781
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5253485899678468,
      "learning_rate": 8.527367365341409e-05,
      "loss": 2.9453,
      "step": 1782
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5082065429815835,
      "learning_rate": 8.485927991043757e-05,
      "loss": 2.8965,
      "step": 1783
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5306113308273891,
      "learning_rate": 8.444580212271125e-05,
      "loss": 2.9512,
      "step": 1784
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.46876597960871685,
      "learning_rate": 8.403324120252159e-05,
      "loss": 2.8164,
      "step": 1785
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4886158346097214,
      "learning_rate": 8.362159806013175e-05,
      "loss": 2.875,
      "step": 1786
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4959047440984803,
      "learning_rate": 8.321087360377988e-05,
      "loss": 2.9395,
      "step": 1787
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5030816088884061,
      "learning_rate": 8.280106873967752e-05,
      "loss": 2.9453,
      "step": 1788
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5066602833540801,
      "learning_rate": 8.239218437200679e-05,
      "loss": 2.793,
      "step": 1789
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.49216824821019695,
      "learning_rate": 8.198422140291939e-05,
      "loss": 2.8613,
      "step": 1790
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4868128742584423,
      "learning_rate": 8.157718073253351e-05,
      "loss": 2.873,
      "step": 1791
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5333261187539519,
      "learning_rate": 8.117106325893287e-05,
      "loss": 2.8418,
      "step": 1792
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4818651051998806,
      "learning_rate": 8.076586987816404e-05,
      "loss": 2.8516,
      "step": 1793
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.47430545908331434,
      "learning_rate": 8.036160148423449e-05,
      "loss": 2.7793,
      "step": 1794
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5422719147141654,
      "learning_rate": 7.995825896911141e-05,
      "loss": 2.8535,
      "step": 1795
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4971410241072209,
      "learning_rate": 7.955584322271853e-05,
      "loss": 2.9043,
      "step": 1796
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4936675347121298,
      "learning_rate": 7.915435513293523e-05,
      "loss": 2.7715,
      "step": 1797
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.5112184297224387,
      "learning_rate": 7.875379558559387e-05,
      "loss": 2.8965,
      "step": 1798
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.49661494007002555,
      "learning_rate": 7.835416546447838e-05,
      "loss": 2.8906,
      "step": 1799
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.4929307042343585,
      "learning_rate": 7.795546565132167e-05,
      "loss": 2.8125,
      "step": 1800
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5075022291178889,
      "learning_rate": 7.755769702580412e-05,
      "loss": 2.9238,
      "step": 1801
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5223451150652887,
      "learning_rate": 7.716086046555193e-05,
      "loss": 2.875,
      "step": 1802
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.48139694724876625,
      "learning_rate": 7.676495684613432e-05,
      "loss": 2.918,
      "step": 1803
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.47738670850125114,
      "learning_rate": 7.636998704106252e-05,
      "loss": 2.9043,
      "step": 1804
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5321998028961836,
      "learning_rate": 7.597595192178702e-05,
      "loss": 2.8926,
      "step": 1805
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.48488837460630246,
      "learning_rate": 7.558285235769646e-05,
      "loss": 2.8613,
      "step": 1806
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5044370658814428,
      "learning_rate": 7.519068921611494e-05,
      "loss": 2.8887,
      "step": 1807
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5098837606772774,
      "learning_rate": 7.479946336230047e-05,
      "loss": 2.8672,
      "step": 1808
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.4828483981523827,
      "learning_rate": 7.440917565944349e-05,
      "loss": 2.7441,
      "step": 1809
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.4583340175303632,
      "learning_rate": 7.4019826968664e-05,
      "loss": 2.7754,
      "step": 1810
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.45046448266520356,
      "learning_rate": 7.363141814901053e-05,
      "loss": 2.7852,
      "step": 1811
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.4637860330376235,
      "learning_rate": 7.32439500574577e-05,
      "loss": 2.7734,
      "step": 1812
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.509181154714116,
      "learning_rate": 7.285742354890473e-05,
      "loss": 2.8477,
      "step": 1813
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.4805390678559622,
      "learning_rate": 7.247183947617325e-05,
      "loss": 2.9199,
      "step": 1814
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.4912545711981879,
      "learning_rate": 7.20871986900053e-05,
      "loss": 2.8887,
      "step": 1815
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.4868328463323789,
      "learning_rate": 7.170350203906218e-05,
      "loss": 2.8691,
      "step": 1816
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.472112119993882,
      "learning_rate": 7.132075036992158e-05,
      "loss": 2.8145,
      "step": 1817
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5307751065959144,
      "learning_rate": 7.093894452707666e-05,
      "loss": 2.8457,
      "step": 1818
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.5020124208136112,
      "learning_rate": 7.055808535293334e-05,
      "loss": 2.8887,
      "step": 1819
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.4707204110806975,
      "learning_rate": 7.017817368780888e-05,
      "loss": 2.8359,
      "step": 1820
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.49825919618364645,
      "learning_rate": 6.979921036993042e-05,
      "loss": 2.7617,
      "step": 1821
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.49845840936246016,
      "learning_rate": 6.942119623543202e-05,
      "loss": 2.9316,
      "step": 1822
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5324845080486044,
      "learning_rate": 6.904413211835414e-05,
      "loss": 2.8301,
      "step": 1823
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.501132092741084,
      "learning_rate": 6.866801885064056e-05,
      "loss": 2.8672,
      "step": 1824
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.4993290027476537,
      "learning_rate": 6.829285726213769e-05,
      "loss": 2.8633,
      "step": 1825
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.4829551684056042,
      "learning_rate": 6.79186481805918e-05,
      "loss": 2.8496,
      "step": 1826
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.48553330156312646,
      "learning_rate": 6.754539243164754e-05,
      "loss": 2.7539,
      "step": 1827
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.4936218411771368,
      "learning_rate": 6.717309083884654e-05,
      "loss": 2.9473,
      "step": 1828
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.47723726185889354,
      "learning_rate": 6.680174422362468e-05,
      "loss": 2.8887,
      "step": 1829
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5014983645256196,
      "learning_rate": 6.643135340531136e-05,
      "loss": 2.877,
      "step": 1830
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.48931829001756666,
      "learning_rate": 6.606191920112664e-05,
      "loss": 2.8555,
      "step": 1831
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5049558030839639,
      "learning_rate": 6.569344242618036e-05,
      "loss": 2.8672,
      "step": 1832
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5258255025915372,
      "learning_rate": 6.532592389346958e-05,
      "loss": 2.8555,
      "step": 1833
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5155302480179365,
      "learning_rate": 6.495936441387713e-05,
      "loss": 2.8555,
      "step": 1834
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.5209127197884013,
      "learning_rate": 6.459376479617013e-05,
      "loss": 2.9219,
      "step": 1835
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.526337420054452,
      "learning_rate": 6.422912584699752e-05,
      "loss": 2.8828,
      "step": 1836
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.4814274644771478,
      "learning_rate": 6.386544837088904e-05,
      "loss": 2.8887,
      "step": 1837
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.4655462629788751,
      "learning_rate": 6.350273317025251e-05,
      "loss": 2.793,
      "step": 1838
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.50291587876603,
      "learning_rate": 6.314098104537324e-05,
      "loss": 2.8301,
      "step": 1839
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.4783000430924539,
      "learning_rate": 6.278019279441122e-05,
      "loss": 2.8965,
      "step": 1840
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.4862548962578856,
      "learning_rate": 6.242036921339972e-05,
      "loss": 2.8027,
      "step": 1841
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.48944375549542335,
      "learning_rate": 6.206151109624402e-05,
      "loss": 2.791,
      "step": 1842
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5592580159641669,
      "learning_rate": 6.170361923471868e-05,
      "loss": 2.8184,
      "step": 1843
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5170461862044188,
      "learning_rate": 6.134669441846691e-05,
      "loss": 2.8613,
      "step": 1844
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.510719794438695,
      "learning_rate": 6.099073743499772e-05,
      "loss": 2.8906,
      "step": 1845
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.525518416884197,
      "learning_rate": 6.063574906968511e-05,
      "loss": 2.9082,
      "step": 1846
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5261947594026023,
      "learning_rate": 6.028173010576582e-05,
      "loss": 2.8496,
      "step": 1847
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.4941297584001343,
      "learning_rate": 5.9928681324337544e-05,
      "loss": 2.6953,
      "step": 1848
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5076158016236095,
      "learning_rate": 5.957660350435773e-05,
      "loss": 2.8945,
      "step": 1849
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.4987757911292849,
      "learning_rate": 5.922549742264122e-05,
      "loss": 2.7148,
      "step": 1850
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.49550571588750053,
      "learning_rate": 5.8875363853859166e-05,
      "loss": 2.875,
      "step": 1851
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.4896004301885852,
      "learning_rate": 5.852620357053651e-05,
      "loss": 2.8301,
      "step": 1852
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.508296109478311,
      "learning_rate": 5.8178017343051336e-05,
      "loss": 2.8691,
      "step": 1853
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5114982293110376,
      "learning_rate": 5.783080593963219e-05,
      "loss": 2.9238,
      "step": 1854
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5125417752497047,
      "learning_rate": 5.748457012635683e-05,
      "loss": 2.791,
      "step": 1855
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5247138583746239,
      "learning_rate": 5.713931066715078e-05,
      "loss": 2.8516,
      "step": 1856
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5347579972795603,
      "learning_rate": 5.679502832378497e-05,
      "loss": 2.8496,
      "step": 1857
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5335173611012883,
      "learning_rate": 5.645172385587482e-05,
      "loss": 2.7949,
      "step": 1858
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5059912936583202,
      "learning_rate": 5.6109398020877834e-05,
      "loss": 2.9316,
      "step": 1859
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5051041944123473,
      "learning_rate": 5.576805157409265e-05,
      "loss": 2.9043,
      "step": 1860
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.49749113710429294,
      "learning_rate": 5.542768526865677e-05,
      "loss": 2.8965,
      "step": 1861
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.4824794027815075,
      "learning_rate": 5.508829985554509e-05,
      "loss": 2.8301,
      "step": 1862
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5193513287222369,
      "learning_rate": 5.474989608356856e-05,
      "loss": 2.9297,
      "step": 1863
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.49563237889993905,
      "learning_rate": 5.441247469937194e-05,
      "loss": 2.8457,
      "step": 1864
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5445044049156266,
      "learning_rate": 5.407603644743286e-05,
      "loss": 2.916,
      "step": 1865
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5359645114774577,
      "learning_rate": 5.374058207005944e-05,
      "loss": 2.8672,
      "step": 1866
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.48375033402664347,
      "learning_rate": 5.3406112307389066e-05,
      "loss": 2.877,
      "step": 1867
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5415999353294437,
      "learning_rate": 5.3072627897386926e-05,
      "loss": 2.957,
      "step": 1868
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5499409403332264,
      "learning_rate": 5.27401295758439e-05,
      "loss": 3.0059,
      "step": 1869
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5343432218081927,
      "learning_rate": 5.2408618076375315e-05,
      "loss": 2.8574,
      "step": 1870
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5143815205222649,
      "learning_rate": 5.207809413041914e-05,
      "loss": 2.9102,
      "step": 1871
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.48251342426613936,
      "learning_rate": 5.174855846723459e-05,
      "loss": 2.7246,
      "step": 1872
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5283756705811193,
      "learning_rate": 5.1420011813900104e-05,
      "loss": 2.8945,
      "step": 1873
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5020813630003708,
      "learning_rate": 5.109245489531211e-05,
      "loss": 2.8809,
      "step": 1874
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.4895514129387379,
      "learning_rate": 5.0765888434183446e-05,
      "loss": 2.7988,
      "step": 1875
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5778235142929379,
      "learning_rate": 5.0440313151041364e-05,
      "loss": 2.8574,
      "step": 1876
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.48962124716980465,
      "learning_rate": 5.011572976422657e-05,
      "loss": 2.8242,
      "step": 1877
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5155151802936092,
      "learning_rate": 4.9792138989890825e-05,
      "loss": 2.9688,
      "step": 1878
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.52005452639399,
      "learning_rate": 4.9469541541996234e-05,
      "loss": 2.8613,
      "step": 1879
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.4813393211448455,
      "learning_rate": 4.914793813231305e-05,
      "loss": 2.8184,
      "step": 1880
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5395277913296284,
      "learning_rate": 4.882732947041818e-05,
      "loss": 2.9102,
      "step": 1881
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5014434746232789,
      "learning_rate": 4.850771626369416e-05,
      "loss": 2.834,
      "step": 1882
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5119329319480426,
      "learning_rate": 4.818909921732662e-05,
      "loss": 2.8711,
      "step": 1883
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5121000282585269,
      "learning_rate": 4.787147903430383e-05,
      "loss": 2.8926,
      "step": 1884
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5036195684686482,
      "learning_rate": 4.755485641541424e-05,
      "loss": 2.8535,
      "step": 1885
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.5182190828754045,
      "learning_rate": 4.723923205924557e-05,
      "loss": 2.9121,
      "step": 1886
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.4928992349979799,
      "learning_rate": 4.6924606662182736e-05,
      "loss": 2.8047,
      "step": 1887
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.5114983887709682,
      "learning_rate": 4.6610980918406596e-05,
      "loss": 2.9434,
      "step": 1888
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.4932099772433257,
      "learning_rate": 4.629835551989276e-05,
      "loss": 2.916,
      "step": 1889
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.5527829851891896,
      "learning_rate": 4.5986731156409224e-05,
      "loss": 2.8633,
      "step": 1890
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.5360119202704483,
      "learning_rate": 4.567610851551568e-05,
      "loss": 2.8496,
      "step": 1891
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.48445668584432794,
      "learning_rate": 4.536648828256146e-05,
      "loss": 2.8438,
      "step": 1892
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.4926030536668207,
      "learning_rate": 4.505787114068433e-05,
      "loss": 2.8086,
      "step": 1893
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.4744457781068921,
      "learning_rate": 4.4750257770808764e-05,
      "loss": 2.8535,
      "step": 1894
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.4805794932980813,
      "learning_rate": 4.444364885164448e-05,
      "loss": 2.8984,
      "step": 1895
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.49136182931515254,
      "learning_rate": 4.413804505968533e-05,
      "loss": 2.9258,
      "step": 1896
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.46594343789315734,
      "learning_rate": 4.3833447069206944e-05,
      "loss": 2.8301,
      "step": 1897
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.4855456938411781,
      "learning_rate": 4.352985555226635e-05,
      "loss": 2.8242,
      "step": 1898
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.540688576004524,
      "learning_rate": 4.322727117869951e-05,
      "loss": 2.9043,
      "step": 1899
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.5249263562210276,
      "learning_rate": 4.29256946161205e-05,
      "loss": 2.8711,
      "step": 1900
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.47320006989276275,
      "learning_rate": 4.262512652991968e-05,
      "loss": 2.8438,
      "step": 1901
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.556620865186689,
      "learning_rate": 4.2325567583262113e-05,
      "loss": 2.8574,
      "step": 1902
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.4901305415286684,
      "learning_rate": 4.2027018437086895e-05,
      "loss": 2.8105,
      "step": 1903
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.49587026311349097,
      "learning_rate": 4.172947975010449e-05,
      "loss": 2.7871,
      "step": 1904
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.4699211173713392,
      "learning_rate": 4.143295217879645e-05,
      "loss": 2.8555,
      "step": 1905
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.48272793264883657,
      "learning_rate": 4.113743637741296e-05,
      "loss": 2.7617,
      "step": 1906
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.5338135894201278,
      "learning_rate": 4.084293299797226e-05,
      "loss": 2.8496,
      "step": 1907
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.5280330673875523,
      "learning_rate": 4.054944269025862e-05,
      "loss": 2.7734,
      "step": 1908
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.523125266302465,
      "learning_rate": 4.025696610182095e-05,
      "loss": 2.9395,
      "step": 1909
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5026167411527795,
      "learning_rate": 3.996550387797187e-05,
      "loss": 2.8125,
      "step": 1910
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.49230172939426936,
      "learning_rate": 3.9675056661785556e-05,
      "loss": 2.957,
      "step": 1911
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5153103526197295,
      "learning_rate": 3.9385625094097154e-05,
      "loss": 2.791,
      "step": 1912
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5028648851771881,
      "learning_rate": 3.909720981350034e-05,
      "loss": 2.8887,
      "step": 1913
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5093961969726979,
      "learning_rate": 3.880981145634704e-05,
      "loss": 2.8477,
      "step": 1914
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5594546756545105,
      "learning_rate": 3.852343065674507e-05,
      "loss": 2.9844,
      "step": 1915
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.4980908582178483,
      "learning_rate": 3.8238068046557276e-05,
      "loss": 2.8906,
      "step": 1916
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.497370922057364,
      "learning_rate": 3.795372425540006e-05,
      "loss": 2.8965,
      "step": 1917
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5052361864586828,
      "learning_rate": 3.76703999106418e-05,
      "loss": 2.8516,
      "step": 1918
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5166094051272364,
      "learning_rate": 3.7388095637401754e-05,
      "loss": 2.8945,
      "step": 1919
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5054791749855423,
      "learning_rate": 3.7106812058548376e-05,
      "loss": 2.9512,
      "step": 1920
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.49965435013030585,
      "learning_rate": 3.682654979469807e-05,
      "loss": 2.8516,
      "step": 1921
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5439216585726305,
      "learning_rate": 3.654730946421403e-05,
      "loss": 2.9629,
      "step": 1922
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5058251862083454,
      "learning_rate": 3.6269091683204466e-05,
      "loss": 2.9023,
      "step": 1923
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5243853536975468,
      "learning_rate": 3.5991897065521693e-05,
      "loss": 2.9492,
      "step": 1924
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5010741971292307,
      "learning_rate": 3.571572622276026e-05,
      "loss": 2.8711,
      "step": 1925
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.4830394785420769,
      "learning_rate": 3.544057976425619e-05,
      "loss": 2.8301,
      "step": 1926
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5543325404571208,
      "learning_rate": 3.5166458297085146e-05,
      "loss": 2.873,
      "step": 1927
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.4794439522912151,
      "learning_rate": 3.489336242606111e-05,
      "loss": 2.8828,
      "step": 1928
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5319846267225591,
      "learning_rate": 3.462129275373577e-05,
      "loss": 2.8711,
      "step": 1929
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.4962505504377042,
      "learning_rate": 3.4350249880395924e-05,
      "loss": 2.8984,
      "step": 1930
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5274700527908593,
      "learning_rate": 3.408023440406355e-05,
      "loss": 2.8926,
      "step": 1931
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.48911207659263195,
      "learning_rate": 3.381124692049331e-05,
      "loss": 2.9492,
      "step": 1932
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5094790409014194,
      "learning_rate": 3.354328802317197e-05,
      "loss": 2.8848,
      "step": 1933
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.47815734657298803,
      "learning_rate": 3.327635830331677e-05,
      "loss": 2.8008,
      "step": 1934
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.48888081161864955,
      "learning_rate": 3.3010458349874206e-05,
      "loss": 2.9512,
      "step": 1935
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.4811034504343436,
      "learning_rate": 3.2745588749518775e-05,
      "loss": 2.8613,
      "step": 1936
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.49294449526262923,
      "learning_rate": 3.248175008665161e-05,
      "loss": 2.9102,
      "step": 1937
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5156499470729718,
      "learning_rate": 3.221894294339911e-05,
      "loss": 2.8789,
      "step": 1938
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.48417468970889493,
      "learning_rate": 3.1957167899611836e-05,
      "loss": 2.8906,
      "step": 1939
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.4802202917914823,
      "learning_rate": 3.169642553286334e-05,
      "loss": 2.8496,
      "step": 1940
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.505758079248983,
      "learning_rate": 3.143671641844831e-05,
      "loss": 2.7812,
      "step": 1941
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.49615846099458566,
      "learning_rate": 3.117804112938205e-05,
      "loss": 2.7461,
      "step": 1942
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.48568567317089967,
      "learning_rate": 3.092040023639869e-05,
      "loss": 2.8359,
      "step": 1943
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.483287930621245,
      "learning_rate": 3.066379430795002e-05,
      "loss": 2.8164,
      "step": 1944
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.48372272546958384,
      "learning_rate": 3.040822391020459e-05,
      "loss": 2.7832,
      "step": 1945
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.4905067284897237,
      "learning_rate": 3.0153689607045842e-05,
      "loss": 2.877,
      "step": 1946
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.49246259839894646,
      "learning_rate": 2.9900191960071545e-05,
      "loss": 2.8809,
      "step": 1947
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5228460461222556,
      "learning_rate": 2.9647731528591848e-05,
      "loss": 2.875,
      "step": 1948
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5319839767087139,
      "learning_rate": 2.9396308869628795e-05,
      "loss": 2.8359,
      "step": 1949
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5378102703698687,
      "learning_rate": 2.914592453791448e-05,
      "loss": 2.8789,
      "step": 1950
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5095677066943244,
      "learning_rate": 2.8896579085889994e-05,
      "loss": 2.9551,
      "step": 1951
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4808899650251221,
      "learning_rate": 2.86482730637046e-05,
      "loss": 2.9355,
      "step": 1952
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.529544600575308,
      "learning_rate": 2.840100701921383e-05,
      "loss": 2.8262,
      "step": 1953
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4792007935477995,
      "learning_rate": 2.8154781497978898e-05,
      "loss": 2.9219,
      "step": 1954
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5170443594375739,
      "learning_rate": 2.7909597043265013e-05,
      "loss": 2.8984,
      "step": 1955
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5035885804794545,
      "learning_rate": 2.7665454196040662e-05,
      "loss": 2.6934,
      "step": 1956
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5120111273835907,
      "learning_rate": 2.7422353494975905e-05,
      "loss": 2.8418,
      "step": 1957
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.509221868634404,
      "learning_rate": 2.7180295476441573e-05,
      "loss": 2.7988,
      "step": 1958
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4816691154277228,
      "learning_rate": 2.6939280674508016e-05,
      "loss": 2.8926,
      "step": 1959
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5238688442217041,
      "learning_rate": 2.669930962094358e-05,
      "loss": 2.8105,
      "step": 1960
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4878858939514403,
      "learning_rate": 2.6460382845214126e-05,
      "loss": 2.9512,
      "step": 1961
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5204374468649152,
      "learning_rate": 2.6222500874481025e-05,
      "loss": 2.8125,
      "step": 1962
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5302850917814073,
      "learning_rate": 2.5985664233600827e-05,
      "loss": 2.8984,
      "step": 1963
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5213953914584617,
      "learning_rate": 2.574987344512336e-05,
      "loss": 2.8887,
      "step": 1964
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5254476282663177,
      "learning_rate": 2.5515129029290984e-05,
      "loss": 2.8105,
      "step": 1965
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4898564293967553,
      "learning_rate": 2.5281431504037556e-05,
      "loss": 2.8711,
      "step": 1966
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5111678832016393,
      "learning_rate": 2.504878138498684e-05,
      "loss": 2.8457,
      "step": 1967
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5109690272493516,
      "learning_rate": 2.48171791854519e-05,
      "loss": 2.9961,
      "step": 1968
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5224024212345616,
      "learning_rate": 2.4586625416433473e-05,
      "loss": 2.9121,
      "step": 1969
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5382583790453896,
      "learning_rate": 2.435712058661921e-05,
      "loss": 2.877,
      "step": 1970
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5159511440618805,
      "learning_rate": 2.4128665202382327e-05,
      "loss": 2.8516,
      "step": 1971
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5168640611984879,
      "learning_rate": 2.3901259767780515e-05,
      "loss": 2.9043,
      "step": 1972
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.505945167287823,
      "learning_rate": 2.367490478455514e-05,
      "loss": 2.8555,
      "step": 1973
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.4952191004542427,
      "learning_rate": 2.3449600752129597e-05,
      "loss": 2.8457,
      "step": 1974
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.4847311228026852,
      "learning_rate": 2.3225348167608685e-05,
      "loss": 2.8711,
      "step": 1975
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.49122814060416853,
      "learning_rate": 2.3002147525777118e-05,
      "loss": 2.793,
      "step": 1976
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5114297011787718,
      "learning_rate": 2.2779999319098856e-05,
      "loss": 2.9277,
      "step": 1977
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.4834967803983911,
      "learning_rate": 2.255890403771571e-05,
      "loss": 2.8613,
      "step": 1978
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5331507821294307,
      "learning_rate": 2.233886216944614e-05,
      "loss": 2.8105,
      "step": 1979
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5121142663174697,
      "learning_rate": 2.211987419978484e-05,
      "loss": 2.875,
      "step": 1980
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5252565035788415,
      "learning_rate": 2.1901940611900705e-05,
      "loss": 2.9766,
      "step": 1981
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5561441252082767,
      "learning_rate": 2.168506188663666e-05,
      "loss": 2.8301,
      "step": 1982
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5086373395930184,
      "learning_rate": 2.1469238502507925e-05,
      "loss": 2.8887,
      "step": 1983
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5346558170483621,
      "learning_rate": 2.125447093570154e-05,
      "loss": 2.875,
      "step": 1984
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.4921368109422058,
      "learning_rate": 2.1040759660074793e-05,
      "loss": 2.7871,
      "step": 1985
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5374462351213145,
      "learning_rate": 2.0828105147154273e-05,
      "loss": 2.8906,
      "step": 1986
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5252162223329837,
      "learning_rate": 2.061650786613545e-05,
      "loss": 2.8809,
      "step": 1987
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5159050913538458,
      "learning_rate": 2.040596828388058e-05,
      "loss": 2.8984,
      "step": 1988
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5380797355522905,
      "learning_rate": 2.019648686491865e-05,
      "loss": 2.8145,
      "step": 1989
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5075948115940392,
      "learning_rate": 1.9988064071443767e-05,
      "loss": 2.7754,
      "step": 1990
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5094360370506645,
      "learning_rate": 1.9780700363314253e-05,
      "loss": 2.9336,
      "step": 1991
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.4944720579424186,
      "learning_rate": 1.957439619805196e-05,
      "loss": 2.7812,
      "step": 1992
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.47168422650143543,
      "learning_rate": 1.9369152030840554e-05,
      "loss": 2.8672,
      "step": 1993
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5367906646800584,
      "learning_rate": 1.916496831452552e-05,
      "loss": 2.918,
      "step": 1994
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.49668486267386164,
      "learning_rate": 1.8961845499611998e-05,
      "loss": 2.8887,
      "step": 1995
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5111467969921034,
      "learning_rate": 1.8759784034264925e-05,
      "loss": 2.7656,
      "step": 1996
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.4933277561909301,
      "learning_rate": 1.855878436430708e-05,
      "loss": 2.8457,
      "step": 1997
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5150920340628754,
      "learning_rate": 1.835884693321871e-05,
      "loss": 2.8574,
      "step": 1998
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.4873301035895728,
      "learning_rate": 1.8159972182136386e-05,
      "loss": 2.7969,
      "step": 1999
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.522029283501159,
      "learning_rate": 1.7962160549851945e-05,
      "loss": 2.875,
      "step": 2000
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5393772586283326,
      "learning_rate": 1.776541247281177e-05,
      "loss": 2.7793,
      "step": 2001
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.4949211566294339,
      "learning_rate": 1.7569728385115224e-05,
      "loss": 2.8496,
      "step": 2002
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.49926809334786654,
      "learning_rate": 1.7375108718514665e-05,
      "loss": 2.9512,
      "step": 2003
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.47236474404084866,
      "learning_rate": 1.7181553902413438e-05,
      "loss": 2.8418,
      "step": 2004
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.48638329931417124,
      "learning_rate": 1.698906436386577e-05,
      "loss": 2.7812,
      "step": 2005
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5644851016768043,
      "learning_rate": 1.679764052757532e-05,
      "loss": 2.8594,
      "step": 2006
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5002794560758699,
      "learning_rate": 1.6607282815894464e-05,
      "loss": 3.0176,
      "step": 2007
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5218242876776996,
      "learning_rate": 1.6417991648823405e-05,
      "loss": 2.8965,
      "step": 2008
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5270784277942508,
      "learning_rate": 1.6229767444008835e-05,
      "loss": 2.8848,
      "step": 2009
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.49530106498601684,
      "learning_rate": 1.604261061674378e-05,
      "loss": 2.8652,
      "step": 2010
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5131780475957382,
      "learning_rate": 1.5856521579965865e-05,
      "loss": 2.8965,
      "step": 2011
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5063089656670277,
      "learning_rate": 1.5671500744256938e-05,
      "loss": 2.9004,
      "step": 2012
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.4886967653241343,
      "learning_rate": 1.5487548517841953e-05,
      "loss": 2.8613,
      "step": 2013
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5026751643276293,
      "learning_rate": 1.530466530658814e-05,
      "loss": 2.916,
      "step": 2014
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.515028365432411,
      "learning_rate": 1.5122851514004054e-05,
      "loss": 2.8223,
      "step": 2015
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.4676462046592939,
      "learning_rate": 1.4942107541238703e-05,
      "loss": 2.8438,
      "step": 2016
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5011297653333389,
      "learning_rate": 1.4762433787080809e-05,
      "loss": 2.873,
      "step": 2017
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5085514198843581,
      "learning_rate": 1.4583830647957541e-05,
      "loss": 2.8887,
      "step": 2018
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5195455227635553,
      "learning_rate": 1.4406298517934068e-05,
      "loss": 2.877,
      "step": 2019
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5136693508617406,
      "learning_rate": 1.4229837788712562e-05,
      "loss": 2.9258,
      "step": 2020
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5035699067938926,
      "learning_rate": 1.4054448849631085e-05,
      "loss": 2.7832,
      "step": 2021
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5315552258247844,
      "learning_rate": 1.3880132087663145e-05,
      "loss": 2.8691,
      "step": 2022
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.4919069049180412,
      "learning_rate": 1.3706887887416419e-05,
      "loss": 2.9121,
      "step": 2023
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5213897130603874,
      "learning_rate": 1.3534716631132316e-05,
      "loss": 2.8125,
      "step": 2024
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.4968964967467247,
      "learning_rate": 1.3363618698684853e-05,
      "loss": 2.8066,
      "step": 2025
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.4889337873593396,
      "learning_rate": 1.3193594467579728e-05,
      "loss": 2.8711,
      "step": 2026
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.48420981926139495,
      "learning_rate": 1.3024644312954026e-05,
      "loss": 2.873,
      "step": 2027
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5072682759033157,
      "learning_rate": 1.2856768607574564e-05,
      "loss": 2.8359,
      "step": 2028
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5051709638939146,
      "learning_rate": 1.2689967721837947e-05,
      "loss": 2.873,
      "step": 2029
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5141563292552934,
      "learning_rate": 1.2524242023769006e-05,
      "loss": 2.7754,
      "step": 2030
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5441854794140258,
      "learning_rate": 1.2359591879020526e-05,
      "loss": 2.9023,
      "step": 2031
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.4947961927593052,
      "learning_rate": 1.2196017650872081e-05,
      "loss": 2.8965,
      "step": 2032
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5190575983721172,
      "learning_rate": 1.2033519700229367e-05,
      "loss": 2.832,
      "step": 2033
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.49889957003203256,
      "learning_rate": 1.1872098385623586e-05,
      "loss": 2.8203,
      "step": 2034
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.47199726822220045,
      "learning_rate": 1.1711754063210289e-05,
      "loss": 2.8613,
      "step": 2035
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.49514514158317097,
      "learning_rate": 1.155248708676887e-05,
      "loss": 2.8945,
      "step": 2036
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.49357189621670816,
      "learning_rate": 1.1394297807701737e-05,
      "loss": 2.7988,
      "step": 2037
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.4930694859411921,
      "learning_rate": 1.1237186575033254e-05,
      "loss": 2.8516,
      "step": 2038
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.5428211028350911,
      "learning_rate": 1.1081153735409522e-05,
      "loss": 2.8301,
      "step": 2039
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.48946577188737633,
      "learning_rate": 1.0926199633097156e-05,
      "loss": 2.8047,
      "step": 2040
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.47579245705458734,
      "learning_rate": 1.0772324609982787e-05,
      "loss": 2.793,
      "step": 2041
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.47014502373260225,
      "learning_rate": 1.0619529005571893e-05,
      "loss": 2.834,
      "step": 2042
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5237058338276674,
      "learning_rate": 1.0467813156988748e-05,
      "loss": 2.8398,
      "step": 2043
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.47919628098268346,
      "learning_rate": 1.0317177398975031e-05,
      "loss": 2.9375,
      "step": 2044
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.504725517222397,
      "learning_rate": 1.0167622063889326e-05,
      "loss": 2.7695,
      "step": 2045
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5075601025307019,
      "learning_rate": 1.0019147481706625e-05,
      "loss": 2.8672,
      "step": 2046
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5242136331050359,
      "learning_rate": 9.871753980017051e-06,
      "loss": 2.875,
      "step": 2047
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5063374688113211,
      "learning_rate": 9.725441884025855e-06,
      "loss": 2.8184,
      "step": 2048
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5038085238488473,
      "learning_rate": 9.580211516551862e-06,
      "loss": 2.8711,
      "step": 2049
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.4879496375193668,
      "learning_rate": 9.436063198027589e-06,
      "loss": 2.9082,
      "step": 2050
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5065064768583832,
      "learning_rate": 9.292997246497959e-06,
      "loss": 2.9473,
      "step": 2051
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5107348945974086,
      "learning_rate": 9.151013977619693e-06,
      "loss": 2.9277,
      "step": 2052
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.48699038261298233,
      "learning_rate": 9.010113704661038e-06,
      "loss": 2.7852,
      "step": 2053
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5155491324657906,
      "learning_rate": 8.870296738500316e-06,
      "loss": 2.9316,
      "step": 2054
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.48098594550172974,
      "learning_rate": 8.731563387626095e-06,
      "loss": 2.8633,
      "step": 2055
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.4910172382193473,
      "learning_rate": 8.59391395813569e-06,
      "loss": 2.8633,
      "step": 2056
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.45997912661766954,
      "learning_rate": 8.457348753735328e-06,
      "loss": 2.877,
      "step": 2057
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.4870051986514324,
      "learning_rate": 8.321868075738593e-06,
      "loss": 2.8281,
      "step": 2058
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5185583546319343,
      "learning_rate": 8.187472223066371e-06,
      "loss": 2.8887,
      "step": 2059
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.4878624654403821,
      "learning_rate": 8.054161492246136e-06,
      "loss": 2.7383,
      "step": 2060
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5118437518348143,
      "learning_rate": 7.921936177411049e-06,
      "loss": 2.8184,
      "step": 2061
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5060580678749254,
      "learning_rate": 7.790796570299463e-06,
      "loss": 2.8379,
      "step": 2062
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.49754993383336993,
      "learning_rate": 7.660742960254207e-06,
      "loss": 2.8047,
      "step": 2063
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.49682561220610866,
      "learning_rate": 7.531775634222138e-06,
      "loss": 2.8301,
      "step": 2064
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5233614572981364,
      "learning_rate": 7.403894876753192e-06,
      "loss": 2.834,
      "step": 2065
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.4915877369324539,
      "learning_rate": 7.277100970000061e-06,
      "loss": 2.8809,
      "step": 2066
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.48563502375704354,
      "learning_rate": 7.151394193717408e-06,
      "loss": 2.8652,
      "step": 2067
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5399903968872363,
      "learning_rate": 7.026774825261151e-06,
      "loss": 2.9473,
      "step": 2068
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.494167523367013,
      "learning_rate": 6.903243139588233e-06,
      "loss": 2.9238,
      "step": 2069
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5352881048998094,
      "learning_rate": 6.780799409255522e-06,
      "loss": 2.9121,
      "step": 2070
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5418710494307344,
      "learning_rate": 6.659443904419637e-06,
      "loss": 2.9277,
      "step": 2071
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.48041363795349606,
      "learning_rate": 6.539176892836008e-06,
      "loss": 2.9238,
      "step": 2072
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.4816559996192473,
      "learning_rate": 6.4199986398585375e-06,
      "loss": 2.793,
      "step": 2073
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.4724003387433779,
      "learning_rate": 6.3019094084388884e-06,
      "loss": 2.8516,
      "step": 2074
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5019129638942265,
      "learning_rate": 6.18490945912592e-06,
      "loss": 2.8691,
      "step": 2075
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.47816100129634176,
      "learning_rate": 6.068999050065249e-06,
      "loss": 2.8477,
      "step": 2076
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.4952001273525861,
      "learning_rate": 5.9541784369983586e-06,
      "loss": 2.7539,
      "step": 2077
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.4962196181915607,
      "learning_rate": 5.840447873262433e-06,
      "loss": 2.7988,
      "step": 2078
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5107669980430641,
      "learning_rate": 5.727807609789471e-06,
      "loss": 2.8008,
      "step": 2079
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5567347497686707,
      "learning_rate": 5.616257895105892e-06,
      "loss": 2.8887,
      "step": 2080
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.544917732621671,
      "learning_rate": 5.505798975331933e-06,
      "loss": 2.9004,
      "step": 2081
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.49739370477721917,
      "learning_rate": 5.396431094181198e-06,
      "loss": 2.8672,
      "step": 2082
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5158372325073919,
      "learning_rate": 5.288154492960107e-06,
      "loss": 2.9199,
      "step": 2083
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.48661615001062697,
      "learning_rate": 5.1809694105671155e-06,
      "loss": 2.8496,
      "step": 2084
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.4988653482894666,
      "learning_rate": 5.074876083492441e-06,
      "loss": 2.7988,
      "step": 2085
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5292376462786569,
      "learning_rate": 4.96987474581767e-06,
      "loss": 2.791,
      "step": 2086
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5054822787639952,
      "learning_rate": 4.865965629214819e-06,
      "loss": 2.8906,
      "step": 2087
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5079956970492052,
      "learning_rate": 4.763148962946218e-06,
      "loss": 2.7617,
      "step": 2088
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5141018906819552,
      "learning_rate": 4.661424973863681e-06,
      "loss": 2.8262,
      "step": 2089
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.572053124738819,
      "learning_rate": 4.560793886408398e-06,
      "loss": 2.9336,
      "step": 2090
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5084140866418232,
      "learning_rate": 4.461255922609986e-06,
      "loss": 2.875,
      "step": 2091
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5021957701170263,
      "learning_rate": 4.362811302086267e-06,
      "loss": 2.8809,
      "step": 2092
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.48452962089391255,
      "learning_rate": 4.265460242042885e-06,
      "loss": 2.7363,
      "step": 2093
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5033448164820412,
      "learning_rate": 4.169202957272522e-06,
      "loss": 2.8809,
      "step": 2094
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5010299283552588,
      "learning_rate": 4.074039660154738e-06,
      "loss": 2.8809,
      "step": 2095
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.49924643050293965,
      "learning_rate": 3.9799705606551325e-06,
      "loss": 2.8242,
      "step": 2096
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5533819408576396,
      "learning_rate": 3.886995866325294e-06,
      "loss": 2.916,
      "step": 2097
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.49364019474716037,
      "learning_rate": 3.795115782302072e-06,
      "loss": 2.8145,
      "step": 2098
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.4811417167571818,
      "learning_rate": 3.704330511307197e-06,
      "loss": 2.8359,
      "step": 2099
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.487653333475326,
      "learning_rate": 3.614640253646828e-06,
      "loss": 2.7773,
      "step": 2100
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.537195740063443,
      "learning_rate": 3.5260452072110594e-06,
      "loss": 2.8418,
      "step": 2101
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5094443038805662,
      "learning_rate": 3.4385455674737498e-06,
      "loss": 2.8242,
      "step": 2102
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.4996830931414429,
      "learning_rate": 3.3521415274915256e-06,
      "loss": 2.9004,
      "step": 2103
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5139506834902382,
      "learning_rate": 3.2668332779041133e-06,
      "loss": 2.8711,
      "step": 2104
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.4868930235759473,
      "learning_rate": 3.1826210069332838e-06,
      "loss": 2.8672,
      "step": 2105
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5028949543269121,
      "learning_rate": 3.0995049003826324e-06,
      "loss": 2.877,
      "step": 2106
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.47960797369110364,
      "learning_rate": 3.017485141637355e-06,
      "loss": 2.8027,
      "step": 2107
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.4906276597883707,
      "learning_rate": 2.9365619116636376e-06,
      "loss": 2.8516,
      "step": 2108
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5229240710514563,
      "learning_rate": 2.856735389008269e-06,
      "loss": 2.9277,
      "step": 2109
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.4986832059235053,
      "learning_rate": 2.778005749798307e-06,
      "loss": 2.8438,
      "step": 2110
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5253360768278096,
      "learning_rate": 2.700373167740744e-06,
      "loss": 2.8301,
      "step": 2111
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.4900889871223868,
      "learning_rate": 2.62383781412201e-06,
      "loss": 2.7207,
      "step": 2112
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.4958205648625644,
      "learning_rate": 2.5483998578076373e-06,
      "loss": 2.8828,
      "step": 2113
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.48415452403482356,
      "learning_rate": 2.4740594652418736e-06,
      "loss": 2.875,
      "step": 2114
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.49167549330934346,
      "learning_rate": 2.4008168004472917e-06,
      "loss": 2.8438,
      "step": 2115
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5381158968162637,
      "learning_rate": 2.3286720250246253e-06,
      "loss": 2.9102,
      "step": 2116
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.4927585155040314,
      "learning_rate": 2.2576252981520994e-06,
      "loss": 2.8691,
      "step": 2117
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.49963732606794803,
      "learning_rate": 2.1876767765853233e-06,
      "loss": 2.8145,
      "step": 2118
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5036038697273448,
      "learning_rate": 2.118826614656788e-06,
      "loss": 2.8027,
      "step": 2119
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5040441017990028,
      "learning_rate": 2.051074964275701e-06,
      "loss": 2.8301,
      "step": 2120
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5109284357779263,
      "learning_rate": 1.984421974927375e-06,
      "loss": 2.8184,
      "step": 2121
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.47703618319635593,
      "learning_rate": 1.9188677936731734e-06,
      "loss": 2.832,
      "step": 2122
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.4767618373739439,
      "learning_rate": 1.8544125651501208e-06,
      "loss": 2.8457,
      "step": 2123
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5094607709566231,
      "learning_rate": 1.7910564315704035e-06,
      "loss": 2.7461,
      "step": 2124
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.5154651235153571,
      "learning_rate": 1.7287995327214257e-06,
      "loss": 2.9336,
      "step": 2125
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.48936914782240976,
      "learning_rate": 1.6676420059649754e-06,
      "loss": 2.9062,
      "step": 2126
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.4902688391651598,
      "learning_rate": 1.6075839862374485e-06,
      "loss": 2.7891,
      "step": 2127
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5092520593520721,
      "learning_rate": 1.5486256060492366e-06,
      "loss": 2.959,
      "step": 2128
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.48008389737445284,
      "learning_rate": 1.4907669954844495e-06,
      "loss": 2.834,
      "step": 2129
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5216695563548525,
      "learning_rate": 1.434008282200805e-06,
      "loss": 2.8984,
      "step": 2130
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.48014246160303065,
      "learning_rate": 1.3783495914291844e-06,
      "loss": 2.7988,
      "step": 2131
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5102757850774902,
      "learning_rate": 1.3237910459734104e-06,
      "loss": 2.8887,
      "step": 2132
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.4996987402147346,
      "learning_rate": 1.270332766210025e-06,
      "loss": 2.8379,
      "step": 2133
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5494619205382897,
      "learning_rate": 1.2179748700879012e-06,
      "loss": 2.8848,
      "step": 2134
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5231405780930206,
      "learning_rate": 1.1667174731280205e-06,
      "loss": 2.8438,
      "step": 2135
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.49766374903825383,
      "learning_rate": 1.1165606884234182e-06,
      "loss": 2.8301,
      "step": 2136
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.49397515541550857,
      "learning_rate": 1.0675046266386268e-06,
      "loss": 2.8652,
      "step": 2137
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5555457974942498,
      "learning_rate": 1.019549396009567e-06,
      "loss": 2.8867,
      "step": 2138
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5073213190531907,
      "learning_rate": 9.726951023434348e-07,
      "loss": 2.9082,
      "step": 2139
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5487436893740515,
      "learning_rate": 9.269418490182591e-07,
      "loss": 2.9004,
      "step": 2140
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5226709304802006,
      "learning_rate": 8.822897369827332e-07,
      "loss": 2.8613,
      "step": 2141
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5240907649830736,
      "learning_rate": 8.387388647561611e-07,
      "loss": 2.9082,
      "step": 2142
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5342775931590465,
      "learning_rate": 7.962893284279016e-07,
      "loss": 2.957,
      "step": 2143
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.49991021619706993,
      "learning_rate": 7.549412216574791e-07,
      "loss": 2.9062,
      "step": 2144
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5143717550477217,
      "learning_rate": 7.146946356743067e-07,
      "loss": 2.8438,
      "step": 2145
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.49447580246974077,
      "learning_rate": 6.755496592773524e-07,
      "loss": 2.8027,
      "step": 2146
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.5026406115331522,
      "learning_rate": 6.375063788349733e-07,
      "loss": 2.8223,
      "step": 2147
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.49156213062025284,
      "learning_rate": 6.005648782848594e-07,
      "loss": 2.7109,
      "step": 2148
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5065681811653711,
      "learning_rate": 5.647252391337565e-07,
      "loss": 2.8926,
      "step": 2149
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5151462238234852,
      "learning_rate": 5.299875404572441e-07,
      "loss": 2.8125,
      "step": 2150
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5295968756137654,
      "learning_rate": 4.963518588996796e-07,
      "loss": 2.9336,
      "step": 2151
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.501872659179758,
      "learning_rate": 4.638182686738657e-07,
      "loss": 2.9199,
      "step": 2152
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.510178683693264,
      "learning_rate": 4.3238684156110543e-07,
      "loss": 2.8555,
      "step": 2153
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5016401266845001,
      "learning_rate": 4.020576469108139e-07,
      "loss": 2.8262,
      "step": 2154
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5108980232335145,
      "learning_rate": 3.7283075164046274e-07,
      "loss": 2.8398,
      "step": 2155
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.48924584594790516,
      "learning_rate": 3.4470622023557995e-07,
      "loss": 2.8301,
      "step": 2156
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5075952833604959,
      "learning_rate": 3.176841147492504e-07,
      "loss": 2.7344,
      "step": 2157
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5345009471045464,
      "learning_rate": 2.9176449480244895e-07,
      "loss": 2.8516,
      "step": 2158
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.4924734159169673,
      "learning_rate": 2.6694741758342967e-07,
      "loss": 2.8555,
      "step": 2159
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.4883312990943159,
      "learning_rate": 2.432329378478926e-07,
      "loss": 2.8418,
      "step": 2160
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5012455201078854,
      "learning_rate": 2.2062110791892798e-07,
      "loss": 2.8457,
      "step": 2161
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.51179398910497,
      "learning_rate": 1.9911197768662792e-07,
      "loss": 2.9453,
      "step": 2162
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5001068705333963,
      "learning_rate": 1.7870559460814173e-07,
      "loss": 2.8867,
      "step": 2163
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.49565656952308296,
      "learning_rate": 1.5940200370750947e-07,
      "loss": 2.7207,
      "step": 2164
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5040891636137823,
      "learning_rate": 1.4120124757577291e-07,
      "loss": 2.8613,
      "step": 2165
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5141548134837735,
      "learning_rate": 1.2410336637047603e-07,
      "loss": 2.9316,
      "step": 2166
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5724906799250317,
      "learning_rate": 1.081083978159425e-07,
      "loss": 2.8008,
      "step": 2167
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5009369578889554,
      "learning_rate": 9.321637720310915e-08,
      "loss": 2.8613,
      "step": 2168
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5193523798131139,
      "learning_rate": 7.942733738924845e-08,
      "loss": 2.8945,
      "step": 2169
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.49536993429703996,
      "learning_rate": 6.6741308798135e-08,
      "loss": 2.9512,
      "step": 2170
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5026888972296281,
      "learning_rate": 5.5158319419934546e-08,
      "loss": 2.7891,
      "step": 2171
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5360428290774435,
      "learning_rate": 4.4678394810981906e-08,
      "loss": 2.8242,
      "step": 2172
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.4945412643887671,
      "learning_rate": 3.5301558093947527e-08,
      "loss": 2.9238,
      "step": 2173
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5167544310353913,
      "learning_rate": 2.7027829957559925e-08,
      "loss": 2.8613,
      "step": 2174
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5176650098646315,
      "learning_rate": 1.985722865682771e-08,
      "loss": 2.8867,
      "step": 2175
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5177894110982965,
      "learning_rate": 1.3789770012762048e-08,
      "loss": 2.7852,
      "step": 2176
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5109847315552817,
      "learning_rate": 8.825467412376665e-09,
      "loss": 2.7891,
      "step": 2177
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5206469332070892,
      "learning_rate": 4.96433180879885e-09,
      "loss": 2.8594,
      "step": 2178
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5322490819707721,
      "learning_rate": 2.206371721158451e-09,
      "loss": 2.9883,
      "step": 2179
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5249141222719024,
      "learning_rate": 5.515932345323549e-10,
      "loss": 2.8496,
      "step": 2180
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5015648661839242,
      "learning_rate": 0.0,
      "loss": 2.8906,
      "step": 2181
    },
    {
      "epoch": 1.0,
      "step": 2181,
      "total_flos": 7.084547721092137e+17,
      "train_loss": 2.9983594108207243,
      "train_runtime": 64537.0731,
      "train_samples_per_second": 8.648,
      "train_steps_per_second": 0.034
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 2181,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 7.084547721092137e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}