{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998853342506593,
  "eval_steps": 500,
  "global_step": 2180,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 49.05611038208008,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 6.6089,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 61.09135055541992,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 6.7085,
      "step": 2
    },
    {
      "epoch": 0.0,
      "grad_norm": 43.82002258300781,
      "learning_rate": 4.545454545454546e-05,
      "loss": 6.4985,
      "step": 3
    },
    {
      "epoch": 0.0,
      "grad_norm": 43.28299331665039,
      "learning_rate": 6.060606060606061e-05,
      "loss": 6.1612,
      "step": 4
    },
    {
      "epoch": 0.0,
      "grad_norm": 29.271106719970703,
      "learning_rate": 7.575757575757576e-05,
      "loss": 5.5483,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 19.42603302001953,
      "learning_rate": 9.090909090909092e-05,
      "loss": 5.4081,
      "step": 6
    },
    {
      "epoch": 0.0,
      "grad_norm": 14.389080047607422,
      "learning_rate": 0.00010606060606060606,
      "loss": 5.1234,
      "step": 7
    },
    {
      "epoch": 0.0,
      "grad_norm": 7.732672691345215,
      "learning_rate": 0.00012121212121212122,
      "loss": 4.9549,
      "step": 8
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.11053466796875,
      "learning_rate": 0.00013636363636363637,
      "loss": 4.8459,
      "step": 9
    },
    {
      "epoch": 0.0,
      "grad_norm": 12.35612964630127,
      "learning_rate": 0.00015151515151515152,
      "loss": 4.6515,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.513789176940918,
      "learning_rate": 0.00016666666666666666,
      "loss": 4.627,
      "step": 11
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.0814595222473145,
      "learning_rate": 0.00018181818181818183,
      "loss": 4.5098,
      "step": 12
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.123250007629395,
      "learning_rate": 0.00019696969696969695,
      "loss": 4.5219,
      "step": 13
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.8690412044525146,
      "learning_rate": 0.00021212121212121213,
      "loss": 4.3847,
      "step": 14
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.438519477844238,
      "learning_rate": 0.00022727272727272727,
      "loss": 4.2921,
      "step": 15
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.469417095184326,
      "learning_rate": 0.00024242424242424245,
      "loss": 4.4507,
      "step": 16
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.32462215423584,
      "learning_rate": 0.00025757575757575756,
      "loss": 4.7233,
      "step": 17
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.6508607864379883,
      "learning_rate": 0.00027272727272727274,
      "loss": 4.5556,
      "step": 18
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7355406284332275,
      "learning_rate": 0.0002878787878787879,
      "loss": 4.3456,
      "step": 19
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.39385986328125,
      "learning_rate": 0.00030303030303030303,
      "loss": 4.2825,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3164482116699219,
      "learning_rate": 0.0003181818181818182,
      "loss": 4.2417,
      "step": 21
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3747220039367676,
      "learning_rate": 0.0003333333333333333,
      "loss": 4.0843,
      "step": 22
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.844870567321777,
      "learning_rate": 0.0003484848484848485,
      "loss": 3.9771,
      "step": 23
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.067422389984131,
      "learning_rate": 0.00036363636363636367,
      "loss": 4.3264,
      "step": 24
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.9242212772369385,
      "learning_rate": 0.0003787878787878788,
      "loss": 4.1081,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7006816864013672,
      "learning_rate": 0.0003939393939393939,
      "loss": 4.0386,
      "step": 26
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6274082660675049,
      "learning_rate": 0.00040909090909090913,
      "loss": 4.2164,
      "step": 27
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3575066328048706,
      "learning_rate": 0.00042424242424242425,
      "loss": 4.2402,
      "step": 28
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1377639770507812,
      "learning_rate": 0.0004393939393939394,
      "loss": 3.9349,
      "step": 29
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.031458854675293,
      "learning_rate": 0.00045454545454545455,
      "loss": 4.1607,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.0351365804672241,
      "learning_rate": 0.0004696969696969697,
      "loss": 3.9887,
      "step": 31
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.9311360120773315,
      "learning_rate": 0.0004848484848484849,
      "loss": 4.0263,
      "step": 32
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6516402959823608,
      "learning_rate": 0.0005,
      "loss": 3.946,
      "step": 33
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2509255409240723,
      "learning_rate": 0.0005151515151515151,
      "loss": 4.0039,
      "step": 34
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.47509077191352844,
      "learning_rate": 0.0005303030303030302,
      "loss": 3.9082,
      "step": 35
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.5748637914657593,
      "learning_rate": 0.0005454545454545455,
      "loss": 3.8416,
      "step": 36
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.5836559534072876,
      "learning_rate": 0.0005606060606060606,
      "loss": 4.0142,
      "step": 37
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.48631808161735535,
      "learning_rate": 0.0005757575757575758,
      "loss": 3.9659,
      "step": 38
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9293414354324341,
      "learning_rate": 0.0005909090909090909,
      "loss": 3.9864,
      "step": 39
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.49936798214912415,
      "learning_rate": 0.0006060606060606061,
      "loss": 4.0023,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2671335935592651,
      "learning_rate": 0.0006212121212121212,
      "loss": 3.7964,
      "step": 41
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6419565677642822,
      "learning_rate": 0.0006363636363636364,
      "loss": 3.9522,
      "step": 42
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7484743595123291,
      "learning_rate": 0.0006515151515151515,
      "loss": 3.9049,
      "step": 43
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8421618938446045,
      "learning_rate": 0.0006666666666666666,
      "loss": 3.7719,
      "step": 44
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.3126986622810364,
      "learning_rate": 0.0006818181818181818,
      "loss": 3.7616,
      "step": 45
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.5161375999450684,
      "learning_rate": 0.000696969696969697,
      "loss": 3.8658,
      "step": 46
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3366949558258057,
      "learning_rate": 0.0007121212121212122,
      "loss": 3.8528,
      "step": 47
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.960427761077881,
      "learning_rate": 0.0007272727272727273,
      "loss": 3.9009,
      "step": 48
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.691795825958252,
      "learning_rate": 0.0007424242424242425,
      "loss": 3.953,
      "step": 49
    },
    {
      "epoch": 0.02,
      "grad_norm": 23.268781661987305,
      "learning_rate": 0.0007575757575757576,
      "loss": 4.0239,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.445228099822998,
      "learning_rate": 0.0007727272727272727,
      "loss": 4.0389,
      "step": 51
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.837568759918213,
      "learning_rate": 0.0007878787878787878,
      "loss": 3.9949,
      "step": 52
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.3761377334594727,
      "learning_rate": 0.000803030303030303,
      "loss": 3.9498,
      "step": 53
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.001247406005859,
      "learning_rate": 0.0008181818181818183,
      "loss": 4.074,
      "step": 54
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.599027633666992,
      "learning_rate": 0.0008333333333333334,
      "loss": 4.0778,
      "step": 55
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.450680732727051,
      "learning_rate": 0.0008484848484848485,
      "loss": 4.1507,
      "step": 56
    },
    {
      "epoch": 0.03,
      "grad_norm": 18.08251953125,
      "learning_rate": 0.0008636363636363636,
      "loss": 3.9158,
      "step": 57
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.3970603942871094,
      "learning_rate": 0.0008787878787878789,
      "loss": 3.9554,
      "step": 58
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7382048964500427,
      "learning_rate": 0.000893939393939394,
      "loss": 3.9639,
      "step": 59
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.4568021297454834,
      "learning_rate": 0.0009090909090909091,
      "loss": 3.9046,
      "step": 60
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.6522254943847656,
      "learning_rate": 0.0009242424242424242,
      "loss": 3.9432,
      "step": 61
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.103682518005371,
      "learning_rate": 0.0009393939393939394,
      "loss": 3.9192,
      "step": 62
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.812212944030762,
      "learning_rate": 0.0009545454545454546,
      "loss": 3.9194,
      "step": 63
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.119608879089355,
      "learning_rate": 0.0009696969696969698,
      "loss": 3.944,
      "step": 64
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.544704437255859,
      "learning_rate": 0.000984848484848485,
      "loss": 4.0043,
      "step": 65
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3107653856277466,
      "learning_rate": 0.001,
      "loss": 3.8266,
      "step": 66
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8551742434501648,
      "learning_rate": 0.0009999994478847943,
      "loss": 3.7739,
      "step": 67
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6666598916053772,
      "learning_rate": 0.0009999977915403962,
      "loss": 3.8499,
      "step": 68
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4759408235549927,
      "learning_rate": 0.0009999950309704639,
      "loss": 3.7442,
      "step": 69
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7153065204620361,
      "learning_rate": 0.000999991166181094,
      "loss": 3.8131,
      "step": 70
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0637494325637817,
      "learning_rate": 0.0009999861971808216,
      "loss": 3.7869,
      "step": 71
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.815099835395813,
      "learning_rate": 0.0009999801239806208,
      "loss": 3.7881,
      "step": 72
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7690733671188354,
      "learning_rate": 0.0009999729465939035,
      "loss": 3.6242,
      "step": 73
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5582714080810547,
      "learning_rate": 0.0009999646650365212,
      "loss": 3.7952,
      "step": 74
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6237813234329224,
      "learning_rate": 0.0009999552793267634,
      "loss": 3.6935,
      "step": 75
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5538607239723206,
      "learning_rate": 0.0009999447894853577,
      "loss": 3.7865,
      "step": 76
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5233844518661499,
      "learning_rate": 0.0009999331955354708,
      "loss": 3.8533,
      "step": 77
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9885690212249756,
      "learning_rate": 0.0009999204975027073,
      "loss": 3.5626,
      "step": 78
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.4993220865726471,
      "learning_rate": 0.0009999066954151103,
      "loss": 3.5757,
      "step": 79
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.4215394258499146,
      "learning_rate": 0.0009998917893031614,
      "loss": 3.7416,
      "step": 80
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.765872001647949,
      "learning_rate": 0.0009998757791997801,
      "loss": 3.8413,
      "step": 81
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.4600889682769775,
      "learning_rate": 0.0009998586651403238,
      "loss": 3.762,
      "step": 82
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.078250408172607,
      "learning_rate": 0.0009998404471625885,
      "loss": 3.6755,
      "step": 83
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.358595371246338,
      "learning_rate": 0.0009998211253068078,
      "loss": 3.6621,
      "step": 84
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.9960622787475586,
      "learning_rate": 0.0009998006996156535,
      "loss": 3.7534,
      "step": 85
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.9844720363616943,
      "learning_rate": 0.0009997791701342347,
      "loss": 3.5744,
      "step": 86
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.552330732345581,
      "learning_rate": 0.0009997565369100983,
      "loss": 3.6183,
      "step": 87
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.573028564453125,
      "learning_rate": 0.0009997327999932291,
      "loss": 3.6013,
      "step": 88
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.247642993927002,
      "learning_rate": 0.000999707959436049,
      "loss": 3.6563,
      "step": 89
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8044295310974121,
      "learning_rate": 0.0009996820152934176,
      "loss": 3.6679,
      "step": 90
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.5166529417037964,
      "learning_rate": 0.000999654967622631,
      "loss": 3.6432,
      "step": 91
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1037904024124146,
      "learning_rate": 0.0009996268164834238,
      "loss": 3.5515,
      "step": 92
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2317057847976685,
      "learning_rate": 0.000999597561937966,
      "loss": 3.5183,
      "step": 93
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1288317441940308,
      "learning_rate": 0.0009995672040508656,
      "loss": 3.4872,
      "step": 94
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.6976369023323059,
      "learning_rate": 0.0009995357428891662,
      "loss": 3.4934,
      "step": 95
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.4280134439468384,
      "learning_rate": 0.0009995031785223491,
      "loss": 3.5015,
      "step": 96
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.6591877341270447,
      "learning_rate": 0.000999469511022331,
      "loss": 3.6021,
      "step": 97
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.44199687242507935,
      "learning_rate": 0.0009994347404634657,
      "loss": 3.4375,
      "step": 98
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6339625716209412,
      "learning_rate": 0.0009993988669225423,
      "loss": 3.5664,
      "step": 99
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5039956569671631,
      "learning_rate": 0.000999361890478786,
      "loss": 3.4868,
      "step": 100
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.540084719657898,
      "learning_rate": 0.0009993238112138583,
      "loss": 3.5475,
      "step": 101
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.47040989995002747,
      "learning_rate": 0.0009992846292118554,
      "loss": 3.4011,
      "step": 102
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5364029407501221,
      "learning_rate": 0.000999244344559309,
      "loss": 3.458,
      "step": 103
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.49440041184425354,
      "learning_rate": 0.0009992029573451869,
      "loss": 3.3352,
      "step": 104
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.38074931502342224,
      "learning_rate": 0.0009991604676608905,
      "loss": 3.451,
      "step": 105
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.45962628722190857,
      "learning_rate": 0.0009991168756002568,
      "loss": 3.3855,
      "step": 106
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5590383410453796,
      "learning_rate": 0.0009990721812595574,
      "loss": 3.3664,
      "step": 107
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.4485394060611725,
      "learning_rate": 0.0009990263847374976,
      "loss": 3.3453,
      "step": 108
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.46786612272262573,
      "learning_rate": 0.0009989794861352173,
      "loss": 3.4099,
      "step": 109
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.4511386454105377,
      "learning_rate": 0.0009989314855562905,
      "loss": 3.3186,
      "step": 110
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.4102385938167572,
      "learning_rate": 0.0009988823831067245,
      "loss": 3.2835,
      "step": 111
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.48559245467185974,
      "learning_rate": 0.0009988321788949597,
      "loss": 3.3788,
      "step": 112
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.4731830954551697,
      "learning_rate": 0.0009987808730318709,
      "loss": 3.2736,
      "step": 113
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5001579523086548,
      "learning_rate": 0.0009987284656307644,
      "loss": 3.4078,
      "step": 114
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.42590397596359253,
      "learning_rate": 0.0009986749568073802,
      "loss": 3.2226,
      "step": 115
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6717888116836548,
      "learning_rate": 0.0009986203466798905,
      "loss": 3.2204,
      "step": 116
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6424322724342346,
      "learning_rate": 0.0009985646353688996,
      "loss": 3.2234,
      "step": 117
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.502510130405426,
      "learning_rate": 0.0009985078229974437,
      "loss": 3.3132,
      "step": 118
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.4912613034248352,
      "learning_rate": 0.0009984499096909905,
      "loss": 3.2838,
      "step": 119
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7885680794715881,
      "learning_rate": 0.0009983908955774397,
      "loss": 3.2442,
      "step": 120
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.47041255235671997,
      "learning_rate": 0.0009983307807871211,
      "loss": 3.2154,
      "step": 121
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5945774912834167,
      "learning_rate": 0.0009982695654527965,
      "loss": 3.2614,
      "step": 122
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.45651954412460327,
      "learning_rate": 0.0009982072497096571,
      "loss": 3.2816,
      "step": 123
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.38438332080841064,
      "learning_rate": 0.000998143833695325,
      "loss": 3.2102,
      "step": 124
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4487494230270386,
      "learning_rate": 0.0009980793175498517,
      "loss": 3.1289,
      "step": 125
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4770950675010681,
      "learning_rate": 0.000998013701415719,
      "loss": 3.1223,
      "step": 126
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.44823500514030457,
      "learning_rate": 0.0009979469854378372,
      "loss": 3.1542,
      "step": 127
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.492191880941391,
      "learning_rate": 0.000997879169763546,
      "loss": 3.2323,
      "step": 128
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6692559123039246,
      "learning_rate": 0.000997810254542614,
      "loss": 3.1799,
      "step": 129
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5155183672904968,
      "learning_rate": 0.0009977402399272374,
      "loss": 3.1034,
      "step": 130
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4700222611427307,
      "learning_rate": 0.0009976691260720407,
      "loss": 3.1585,
      "step": 131
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5468966364860535,
      "learning_rate": 0.0009975969131340763,
      "loss": 3.1172,
      "step": 132
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.48506683111190796,
      "learning_rate": 0.0009975236012728236,
      "loss": 3.1301,
      "step": 133
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5923392176628113,
      "learning_rate": 0.0009974491906501886,
      "loss": 3.0943,
      "step": 134
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4653012156486511,
      "learning_rate": 0.0009973736814305049,
      "loss": 3.1011,
      "step": 135
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5904295444488525,
      "learning_rate": 0.0009972970737805312,
      "loss": 3.1572,
      "step": 136
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4462765157222748,
      "learning_rate": 0.0009972193678694525,
      "loss": 3.1185,
      "step": 137
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4942866563796997,
      "learning_rate": 0.0009971405638688794,
      "loss": 3.0149,
      "step": 138
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6523067951202393,
      "learning_rate": 0.0009970606619528475,
      "loss": 3.0513,
      "step": 139
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.580865204334259,
      "learning_rate": 0.000996979662297817,
      "loss": 3.0076,
      "step": 140
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4684850871562958,
      "learning_rate": 0.0009968975650826721,
      "loss": 3.1177,
      "step": 141
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6237471699714661,
      "learning_rate": 0.000996814370488722,
      "loss": 3.0885,
      "step": 142
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5001257061958313,
      "learning_rate": 0.000996730078699698,
      "loss": 3.0325,
      "step": 143
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5620262026786804,
      "learning_rate": 0.0009966446899017558,
      "loss": 3.1361,
      "step": 144
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4912627041339874,
      "learning_rate": 0.0009965582042834728,
      "loss": 3.039,
      "step": 145
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.47537901997566223,
      "learning_rate": 0.0009964706220358492,
      "loss": 3.0361,
      "step": 146
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.41321825981140137,
      "learning_rate": 0.000996381943352307,
      "loss": 2.9145,
      "step": 147
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1040056943893433,
      "learning_rate": 0.0009962921684286896,
      "loss": 3.1141,
      "step": 148
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4951709508895874,
      "learning_rate": 0.0009962012974632614,
      "loss": 2.9915,
      "step": 149
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4407115578651428,
      "learning_rate": 0.0009961093306567075,
      "loss": 3.035,
      "step": 150
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4021231532096863,
      "learning_rate": 0.0009960162682121328,
      "loss": 2.9208,
      "step": 151
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.44193437695503235,
      "learning_rate": 0.0009959221103350623,
      "loss": 3.0084,
      "step": 152
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6703566908836365,
      "learning_rate": 0.0009958268572334394,
      "loss": 3.0833,
      "step": 153
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4026874899864197,
      "learning_rate": 0.0009957305091176274,
      "loss": 3.0112,
      "step": 154
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5607354044914246,
      "learning_rate": 0.0009956330662004075,
      "loss": 2.9948,
      "step": 155
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4443207383155823,
      "learning_rate": 0.0009955345286969779,
      "loss": 3.1082,
      "step": 156
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4193463921546936,
      "learning_rate": 0.0009954348968249551,
      "loss": 2.9863,
      "step": 157
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.43167558312416077,
      "learning_rate": 0.0009953341708043724,
      "loss": 2.913,
      "step": 158
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4508337080478668,
      "learning_rate": 0.0009952323508576793,
      "loss": 2.9389,
      "step": 159
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5412083268165588,
      "learning_rate": 0.0009951294372097406,
      "loss": 2.9692,
      "step": 160
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4407351016998291,
      "learning_rate": 0.0009950254300878378,
      "loss": 2.8958,
      "step": 161
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.43543630838394165,
      "learning_rate": 0.000994920329721666,
      "loss": 2.8977,
      "step": 162
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4715792238712311,
      "learning_rate": 0.0009948141363433356,
      "loss": 2.8346,
      "step": 163
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.486142635345459,
      "learning_rate": 0.00099470685018737,
      "loss": 2.8641,
      "step": 164
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4684126675128937,
      "learning_rate": 0.0009945984714907073,
      "loss": 2.8809,
      "step": 165
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6350711584091187,
      "learning_rate": 0.000994489000492697,
      "loss": 2.9528,
      "step": 166
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.43676528334617615,
      "learning_rate": 0.0009943784374351016,
      "loss": 3.0213,
      "step": 167
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.46802616119384766,
      "learning_rate": 0.0009942667825620951,
      "loss": 2.9303,
      "step": 168
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7225691676139832,
      "learning_rate": 0.0009941540361202634,
      "loss": 3.0007,
      "step": 169
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5254736542701721,
      "learning_rate": 0.0009940401983586022,
      "loss": 2.8792,
      "step": 170
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4368533790111542,
      "learning_rate": 0.000993925269528518,
      "loss": 2.8596,
      "step": 171
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5358726382255554,
      "learning_rate": 0.0009938092498838265,
      "loss": 2.9607,
      "step": 172
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4919986128807068,
      "learning_rate": 0.0009936921396807524,
      "loss": 2.9398,
      "step": 173
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.47951170802116394,
      "learning_rate": 0.0009935739391779292,
      "loss": 2.8017,
      "step": 174
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.479282408952713,
      "learning_rate": 0.000993454648636398,
      "loss": 2.9178,
      "step": 175
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.387382984161377,
      "learning_rate": 0.0009933342683196074,
      "loss": 2.913,
      "step": 176
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5061814188957214,
      "learning_rate": 0.0009932127984934125,
      "loss": 2.8523,
      "step": 177
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4673362970352173,
      "learning_rate": 0.0009930902394260745,
      "loss": 2.9173,
      "step": 178
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5779686570167542,
      "learning_rate": 0.0009929665913882607,
      "loss": 2.917,
      "step": 179
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.45873019099235535,
      "learning_rate": 0.0009928418546530425,
      "loss": 2.8989,
      "step": 180
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5107328295707703,
      "learning_rate": 0.0009927160294958964,
      "loss": 2.8502,
      "step": 181
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4416714310646057,
      "learning_rate": 0.000992589116194702,
      "loss": 2.9178,
      "step": 182
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5338780879974365,
      "learning_rate": 0.000992461115029743,
      "loss": 2.932,
      "step": 183
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7559748888015747,
      "learning_rate": 0.000992332026283704,
      "loss": 2.8565,
      "step": 184
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.45434266328811646,
      "learning_rate": 0.0009922018502416736,
      "loss": 2.8657,
      "step": 185
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.45074620842933655,
      "learning_rate": 0.0009920705871911395,
      "loss": 2.7948,
      "step": 186
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.47285357117652893,
      "learning_rate": 0.0009919382374219915,
      "loss": 2.8778,
      "step": 187
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.43368440866470337,
      "learning_rate": 0.0009918048012265187,
      "loss": 2.7839,
      "step": 188
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5679384469985962,
      "learning_rate": 0.0009916702788994097,
      "loss": 2.9989,
      "step": 189
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5553373098373413,
      "learning_rate": 0.0009915346707377519,
      "loss": 2.8149,
      "step": 190
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4761364459991455,
      "learning_rate": 0.0009913979770410305,
      "loss": 2.7683,
      "step": 191
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6285106539726257,
      "learning_rate": 0.0009912601981111285,
      "loss": 2.7826,
      "step": 192
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5601210594177246,
      "learning_rate": 0.0009911213342523248,
      "loss": 2.8247,
      "step": 193
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4678763151168823,
      "learning_rate": 0.000990981385771295,
      "loss": 2.9561,
      "step": 194
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5190218091011047,
      "learning_rate": 0.00099084035297711,
      "loss": 2.8169,
      "step": 195
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.47038698196411133,
      "learning_rate": 0.000990698236181235,
      "loss": 2.8219,
      "step": 196
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5377885103225708,
      "learning_rate": 0.0009905550356975293,
      "loss": 2.762,
      "step": 197
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.462862104177475,
      "learning_rate": 0.0009904107518422457,
      "loss": 2.7894,
      "step": 198
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4630295932292938,
      "learning_rate": 0.0009902653849340295,
      "loss": 2.7208,
      "step": 199
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6004145741462708,
      "learning_rate": 0.0009901189352939177,
      "loss": 2.8673,
      "step": 200
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4646543562412262,
      "learning_rate": 0.0009899714032453387,
      "loss": 2.8514,
      "step": 201
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5038837790489197,
      "learning_rate": 0.000989822789114111,
      "loss": 2.7585,
      "step": 202
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5100123882293701,
      "learning_rate": 0.0009896730932284434,
      "loss": 2.7767,
      "step": 203
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.49501851201057434,
      "learning_rate": 0.0009895223159189332,
      "loss": 2.6681,
      "step": 204
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5398043394088745,
      "learning_rate": 0.0009893704575185663,
      "loss": 2.8365,
      "step": 205
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.624380350112915,
      "learning_rate": 0.000989217518362716,
      "loss": 2.7743,
      "step": 206
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4865369498729706,
      "learning_rate": 0.0009890634987891425,
      "loss": 2.7299,
      "step": 207
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4835790693759918,
      "learning_rate": 0.0009889083991379917,
      "loss": 2.7662,
      "step": 208
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4626990556716919,
      "learning_rate": 0.0009887522197517954,
      "loss": 2.6468,
      "step": 209
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5346041917800903,
      "learning_rate": 0.0009885949609754693,
      "loss": 2.8021,
      "step": 210
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4837276041507721,
      "learning_rate": 0.000988436623156314,
      "loss": 2.7867,
      "step": 211
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4383073151111603,
      "learning_rate": 0.0009882772066440114,
      "loss": 2.7372,
      "step": 212
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.49998053908348083,
      "learning_rate": 0.0009881167117906276,
      "loss": 2.7446,
      "step": 213
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5193607211112976,
      "learning_rate": 0.0009879551389506084,
      "loss": 2.7158,
      "step": 214
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.47107505798339844,
      "learning_rate": 0.0009877924884807814,
      "loss": 2.7265,
      "step": 215
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5218038558959961,
      "learning_rate": 0.000987628760740354,
      "loss": 2.7205,
      "step": 216
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5073615312576294,
      "learning_rate": 0.0009874639560909118,
      "loss": 2.6802,
      "step": 217
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4863731265068054,
      "learning_rate": 0.0009872980748964202,
      "loss": 2.8071,
      "step": 218
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4944925606250763,
      "learning_rate": 0.000987131117523221,
      "loss": 2.812,
      "step": 219
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.45465075969696045,
      "learning_rate": 0.000986963084340033,
      "loss": 2.7568,
      "step": 220
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5184234976768494,
      "learning_rate": 0.0009867939757179508,
      "loss": 2.7457,
      "step": 221
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.47182419896125793,
      "learning_rate": 0.0009866237920304443,
      "loss": 2.7011,
      "step": 222
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4216427206993103,
      "learning_rate": 0.0009864525336533577,
      "loss": 2.6947,
      "step": 223
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4900585412979126,
      "learning_rate": 0.000986280200964908,
      "loss": 2.7433,
      "step": 224
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7546579837799072,
      "learning_rate": 0.0009861067943456856,
      "loss": 2.6347,
      "step": 225
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4733332097530365,
      "learning_rate": 0.000985932314178652,
      "loss": 2.6229,
      "step": 226
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.49525389075279236,
      "learning_rate": 0.00098575676084914,
      "loss": 2.822,
      "step": 227
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4975442886352539,
      "learning_rate": 0.0009855801347448518,
      "loss": 2.7,
      "step": 228
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.44532665610313416,
      "learning_rate": 0.0009854024362558596,
      "loss": 2.7651,
      "step": 229
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5069783329963684,
      "learning_rate": 0.0009852236657746035,
      "loss": 2.6651,
      "step": 230
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4816335439682007,
      "learning_rate": 0.0009850438236958911,
      "loss": 2.7751,
      "step": 231
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.46806633472442627,
      "learning_rate": 0.0009848629104168966,
      "loss": 2.726,
      "step": 232
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5288969278335571,
      "learning_rate": 0.00098468092633716,
      "loss": 2.6457,
      "step": 233
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5184749960899353,
      "learning_rate": 0.0009844978718585855,
      "loss": 2.7506,
      "step": 234
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4671137034893036,
      "learning_rate": 0.0009843137473854423,
      "loss": 2.6499,
      "step": 235
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4769059717655182,
      "learning_rate": 0.000984128553324362,
      "loss": 2.6436,
      "step": 236
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.44763484597206116,
      "learning_rate": 0.0009839422900843383,
      "loss": 2.6981,
      "step": 237
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4218178391456604,
      "learning_rate": 0.0009837549580767261,
      "loss": 2.7081,
      "step": 238
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.48130500316619873,
      "learning_rate": 0.0009835665577152411,
      "loss": 2.6684,
      "step": 239
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.47011005878448486,
      "learning_rate": 0.000983377089415958,
      "loss": 2.6475,
      "step": 240
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.48468559980392456,
      "learning_rate": 0.0009831865535973102,
      "loss": 2.5658,
      "step": 241
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4610452950000763,
      "learning_rate": 0.0009829949506800885,
      "loss": 2.6623,
      "step": 242
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5319145917892456,
      "learning_rate": 0.0009828022810874405,
      "loss": 2.6645,
      "step": 243
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4960525631904602,
      "learning_rate": 0.0009826085452448693,
      "loss": 2.6483,
      "step": 244
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4172958731651306,
      "learning_rate": 0.000982413743580233,
      "loss": 2.7406,
      "step": 245
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4334375262260437,
      "learning_rate": 0.0009822178765237436,
      "loss": 2.6175,
      "step": 246
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.49076738953590393,
      "learning_rate": 0.0009820209445079654,
      "loss": 2.7278,
      "step": 247
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.3951072692871094,
      "learning_rate": 0.0009818229479678158,
      "loss": 2.6599,
      "step": 248
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.45671579241752625,
      "learning_rate": 0.0009816238873405615,
      "loss": 2.6834,
      "step": 249
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4285607635974884,
      "learning_rate": 0.0009814237630658207,
      "loss": 2.5966,
      "step": 250
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4539140462875366,
      "learning_rate": 0.00098122257558556,
      "loss": 2.5772,
      "step": 251
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.41907447576522827,
      "learning_rate": 0.0009810203253440937,
      "loss": 2.5869,
      "step": 252
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.42088791728019714,
      "learning_rate": 0.0009808170127880837,
      "loss": 2.5804,
      "step": 253
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4299432039260864,
      "learning_rate": 0.000980612638366538,
      "loss": 2.5713,
      "step": 254
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.3915051817893982,
      "learning_rate": 0.0009804072025308096,
      "loss": 2.664,
      "step": 255
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4086713492870331,
      "learning_rate": 0.000980200705734595,
      "loss": 2.6204,
      "step": 256
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6972109079360962,
      "learning_rate": 0.0009799931484339344,
      "loss": 2.6263,
      "step": 257
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4049108624458313,
      "learning_rate": 0.0009797845310872103,
      "loss": 2.5989,
      "step": 258
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.3962024748325348,
      "learning_rate": 0.0009795748541551457,
      "loss": 2.668,
      "step": 259
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.48877549171447754,
      "learning_rate": 0.000979364118100804,
      "loss": 2.548,
      "step": 260
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4326624572277069,
      "learning_rate": 0.0009791523233895875,
      "loss": 2.6921,
      "step": 261
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4732770025730133,
      "learning_rate": 0.0009789394704892364,
      "loss": 2.6307,
      "step": 262
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4196014106273651,
      "learning_rate": 0.0009787255598698282,
      "loss": 2.6187,
      "step": 263
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4075445532798767,
      "learning_rate": 0.0009785105920037758,
      "loss": 2.7217,
      "step": 264
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4759919047355652,
      "learning_rate": 0.0009782945673658275,
      "loss": 2.6488,
      "step": 265
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.42319193482398987,
      "learning_rate": 0.0009780774864330654,
      "loss": 2.6423,
      "step": 266
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.46909773349761963,
      "learning_rate": 0.000977859349684904,
      "loss": 2.614,
      "step": 267
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.40710827708244324,
      "learning_rate": 0.00097764015760309,
      "loss": 2.6718,
      "step": 268
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.3795360326766968,
      "learning_rate": 0.0009774199106717004,
      "loss": 2.5528,
      "step": 269
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4189162850379944,
      "learning_rate": 0.0009771986093771417,
      "loss": 2.5057,
      "step": 270
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.422963410615921,
      "learning_rate": 0.0009769762542081496,
      "loss": 2.52,
      "step": 271
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.39239731431007385,
      "learning_rate": 0.000976752845655786,
      "loss": 2.6365,
      "step": 272
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3975802958011627,
      "learning_rate": 0.0009765283842134411,
      "loss": 2.6005,
      "step": 273
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.40109214186668396,
      "learning_rate": 0.0009763028703768282,
      "loss": 2.649,
      "step": 274
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.39226049184799194,
      "learning_rate": 0.0009760763046439862,
      "loss": 2.5419,
      "step": 275
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3882850408554077,
      "learning_rate": 0.0009758486875152766,
      "loss": 2.5898,
      "step": 276
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3863004744052887,
      "learning_rate": 0.0009756200194933829,
      "loss": 2.6179,
      "step": 277
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.36419129371643066,
      "learning_rate": 0.0009753903010833094,
      "loss": 2.6107,
      "step": 278
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3978222608566284,
      "learning_rate": 0.0009751595327923803,
      "loss": 2.5696,
      "step": 279
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.42029455304145813,
      "learning_rate": 0.0009749277151302382,
      "loss": 2.6684,
      "step": 280
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3953621983528137,
      "learning_rate": 0.0009746948486088435,
      "loss": 2.618,
      "step": 281
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.42553383111953735,
      "learning_rate": 0.0009744609337424727,
      "loss": 2.5721,
      "step": 282
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.41746094822883606,
      "learning_rate": 0.0009742259710477177,
      "loss": 2.582,
      "step": 283
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4230157732963562,
      "learning_rate": 0.0009739899610434841,
      "loss": 2.5395,
      "step": 284
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4207427203655243,
      "learning_rate": 0.0009737529042509913,
      "loss": 2.5807,
      "step": 285
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3665088713169098,
      "learning_rate": 0.0009735148011937693,
      "loss": 2.6044,
      "step": 286
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3642357885837555,
      "learning_rate": 0.00097327565239766,
      "loss": 2.6436,
      "step": 287
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3866119682788849,
      "learning_rate": 0.0009730354583908136,
      "loss": 2.6368,
      "step": 288
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.41564497351646423,
      "learning_rate": 0.0009727942197036895,
      "loss": 2.5336,
      "step": 289
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.37668126821517944,
      "learning_rate": 0.0009725519368690539,
      "loss": 2.581,
      "step": 290
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.39392685890197754,
      "learning_rate": 0.0009723086104219787,
      "loss": 2.5434,
      "step": 291
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4231227934360504,
      "learning_rate": 0.0009720642408998409,
      "loss": 2.5373,
      "step": 292
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.430152952671051,
      "learning_rate": 0.0009718188288423211,
      "loss": 2.5697,
      "step": 293
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3996846675872803,
      "learning_rate": 0.0009715723747914022,
      "loss": 2.5377,
      "step": 294
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3794103264808655,
      "learning_rate": 0.0009713248792913685,
      "loss": 2.5494,
      "step": 295
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3980809152126312,
      "learning_rate": 0.0009710763428888037,
      "loss": 2.5327,
      "step": 296
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.38724425435066223,
      "learning_rate": 0.0009708267661325909,
      "loss": 2.6004,
      "step": 297
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3862708806991577,
      "learning_rate": 0.0009705761495739107,
      "loss": 2.6508,
      "step": 298
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3977988660335541,
      "learning_rate": 0.0009703244937662399,
      "loss": 2.5772,
      "step": 299
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3997924029827118,
      "learning_rate": 0.0009700717992653505,
      "loss": 2.5698,
      "step": 300
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.37248843908309937,
      "learning_rate": 0.0009698180666293083,
      "loss": 2.5195,
      "step": 301
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3771599233150482,
      "learning_rate": 0.000969563296418472,
      "loss": 2.5827,
      "step": 302
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.36208516359329224,
      "learning_rate": 0.0009693074891954914,
      "loss": 2.4728,
      "step": 303
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.38990139961242676,
      "learning_rate": 0.0009690506455253072,
      "loss": 2.5722,
      "step": 304
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4117380678653717,
      "learning_rate": 0.0009687927659751481,
      "loss": 2.6507,
      "step": 305
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3771120309829712,
      "learning_rate": 0.0009685338511145312,
      "loss": 2.5064,
      "step": 306
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3568870425224304,
      "learning_rate": 0.0009682739015152598,
      "loss": 2.6227,
      "step": 307
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.36081087589263916,
      "learning_rate": 0.0009680129177514226,
      "loss": 2.6005,
      "step": 308
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.408766508102417,
      "learning_rate": 0.0009677509003993915,
      "loss": 2.5437,
      "step": 309
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3471553921699524,
      "learning_rate": 0.0009674878500378221,
      "loss": 2.5678,
      "step": 310
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3767811357975006,
      "learning_rate": 0.0009672237672476505,
      "loss": 2.4981,
      "step": 311
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3654491603374481,
      "learning_rate": 0.0009669586526120935,
      "loss": 2.506,
      "step": 312
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.40985193848609924,
      "learning_rate": 0.0009666925067166459,
      "loss": 2.5932,
      "step": 313
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3637598752975464,
      "learning_rate": 0.000966425330149081,
      "loss": 2.576,
      "step": 314
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3899155855178833,
      "learning_rate": 0.0009661571234994475,
      "loss": 2.4626,
      "step": 315
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.3614533543586731,
      "learning_rate": 0.0009658878873600691,
      "loss": 2.4094,
      "step": 316
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.388639897108078,
      "learning_rate": 0.0009656176223255438,
      "loss": 2.5469,
      "step": 317
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.38005900382995605,
      "learning_rate": 0.000965346328992741,
      "loss": 2.6025,
      "step": 318
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.380169153213501,
      "learning_rate": 0.0009650740079608014,
      "loss": 2.5946,
      "step": 319
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.36303073167800903,
      "learning_rate": 0.0009648006598311353,
      "loss": 2.5245,
      "step": 320
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.39976298809051514,
      "learning_rate": 0.0009645262852074214,
      "loss": 2.5162,
      "step": 321
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3849450945854187,
      "learning_rate": 0.0009642508846956053,
      "loss": 2.5225,
      "step": 322
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3813883662223816,
      "learning_rate": 0.0009639744589038983,
      "loss": 2.5552,
      "step": 323
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3495023250579834,
      "learning_rate": 0.0009636970084427759,
      "loss": 2.4896,
      "step": 324
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.38319462537765503,
      "learning_rate": 0.0009634185339249766,
      "loss": 2.6257,
      "step": 325
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3459699749946594,
      "learning_rate": 0.0009631390359655003,
      "loss": 2.5576,
      "step": 326
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.344086617231369,
      "learning_rate": 0.0009628585151816074,
      "loss": 2.5404,
      "step": 327
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3845425546169281,
      "learning_rate": 0.0009625769721928172,
      "loss": 2.5934,
      "step": 328
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.40262702107429504,
      "learning_rate": 0.0009622944076209061,
      "loss": 2.5472,
      "step": 329
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4229961335659027,
      "learning_rate": 0.0009620108220899071,
      "loss": 2.555,
      "step": 330
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.39379259943962097,
      "learning_rate": 0.0009617262162261075,
      "loss": 2.6141,
      "step": 331
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3533181846141815,
      "learning_rate": 0.0009614405906580486,
      "loss": 2.5954,
      "step": 332
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.37115657329559326,
      "learning_rate": 0.000961153946016523,
      "loss": 2.6353,
      "step": 333
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3975081741809845,
      "learning_rate": 0.000960866282934574,
      "loss": 2.5583,
      "step": 334
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.35895800590515137,
      "learning_rate": 0.0009605776020474945,
      "loss": 2.5578,
      "step": 335
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3451128602027893,
      "learning_rate": 0.0009602879039928249,
      "loss": 2.4805,
      "step": 336
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4297550618648529,
      "learning_rate": 0.0009599971894103521,
      "loss": 2.5423,
      "step": 337
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3668675422668457,
      "learning_rate": 0.0009597054589421077,
      "loss": 2.5059,
      "step": 338
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.36110684275627136,
      "learning_rate": 0.0009594127132323669,
      "loss": 2.4485,
      "step": 339
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4115723967552185,
      "learning_rate": 0.0009591189529276474,
      "loss": 2.5983,
      "step": 340
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3835652768611908,
      "learning_rate": 0.0009588241786767072,
      "loss": 2.5086,
      "step": 341
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.35952070355415344,
      "learning_rate": 0.0009585283911305436,
      "loss": 2.5255,
      "step": 342
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3906164765357971,
      "learning_rate": 0.000958231590942392,
      "loss": 2.5548,
      "step": 343
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3656546175479889,
      "learning_rate": 0.0009579337787677238,
      "loss": 2.4467,
      "step": 344
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3494139313697815,
      "learning_rate": 0.0009576349552642456,
      "loss": 2.5981,
      "step": 345
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.377378910779953,
      "learning_rate": 0.0009573351210918975,
      "loss": 2.5726,
      "step": 346
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3847045302391052,
      "learning_rate": 0.0009570342769128514,
      "loss": 2.6193,
      "step": 347
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.40761786699295044,
      "learning_rate": 0.0009567324233915099,
      "loss": 2.5012,
      "step": 348
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.35971599817276,
      "learning_rate": 0.0009564295611945047,
      "loss": 2.5628,
      "step": 349
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.36979958415031433,
      "learning_rate": 0.000956125690990695,
      "loss": 2.5496,
      "step": 350
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.35049283504486084,
      "learning_rate": 0.0009558208134511665,
      "loss": 2.5288,
      "step": 351
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3546333312988281,
      "learning_rate": 0.0009555149292492289,
      "loss": 2.5282,
      "step": 352
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.33911654353141785,
      "learning_rate": 0.0009552080390604159,
      "loss": 2.5346,
      "step": 353
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.38656285405158997,
      "learning_rate": 0.0009549001435624823,
      "loss": 2.5466,
      "step": 354
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.35721737146377563,
      "learning_rate": 0.0009545912434354029,
      "loss": 2.5292,
      "step": 355
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3895017206668854,
      "learning_rate": 0.0009542813393613721,
      "loss": 2.4893,
      "step": 356
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.37827467918395996,
      "learning_rate": 0.0009539704320248006,
      "loss": 2.4437,
      "step": 357
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.37459471821784973,
      "learning_rate": 0.0009536585221123151,
      "loss": 2.5211,
      "step": 358
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.34985047578811646,
      "learning_rate": 0.0009533456103127565,
      "loss": 2.4965,
      "step": 359
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3271530270576477,
      "learning_rate": 0.000953031697317178,
      "loss": 2.535,
      "step": 360
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.33628371357917786,
      "learning_rate": 0.0009527167838188445,
      "loss": 2.4736,
      "step": 361
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.33303433656692505,
      "learning_rate": 0.0009524008705132299,
      "loss": 2.4992,
      "step": 362
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.36088135838508606,
      "learning_rate": 0.0009520839580980166,
      "loss": 2.4855,
      "step": 363
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.37023770809173584,
      "learning_rate": 0.0009517660472730929,
      "loss": 2.4759,
      "step": 364
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3722972571849823,
      "learning_rate": 0.0009514471387405526,
      "loss": 2.4943,
      "step": 365
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3900412619113922,
      "learning_rate": 0.0009511272332046926,
      "loss": 2.5007,
      "step": 366
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3775114417076111,
      "learning_rate": 0.0009508063313720119,
      "loss": 2.5686,
      "step": 367
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.35473060607910156,
      "learning_rate": 0.0009504844339512095,
      "loss": 2.4844,
      "step": 368
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3831462860107422,
      "learning_rate": 0.0009501615416531835,
      "loss": 2.5847,
      "step": 369
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.38184309005737305,
      "learning_rate": 0.0009498376551910285,
      "loss": 2.5477,
      "step": 370
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.36579227447509766,
      "learning_rate": 0.0009495127752800352,
      "loss": 2.4701,
      "step": 371
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3738860487937927,
      "learning_rate": 0.0009491869026376882,
      "loss": 2.4477,
      "step": 372
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.38890451192855835,
      "learning_rate": 0.0009488600379836648,
      "loss": 2.4703,
      "step": 373
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3834860324859619,
      "learning_rate": 0.0009485321820398321,
      "loss": 2.537,
      "step": 374
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3730798661708832,
      "learning_rate": 0.0009482033355302475,
      "loss": 2.6484,
      "step": 375
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.41048023104667664,
      "learning_rate": 0.0009478734991811556,
      "loss": 2.4569,
      "step": 376
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.36914584040641785,
      "learning_rate": 0.0009475426737209871,
      "loss": 2.4538,
      "step": 377
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3763344883918762,
      "learning_rate": 0.000947210859880357,
      "loss": 2.5141,
      "step": 378
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.36605724692344666,
      "learning_rate": 0.0009468780583920631,
      "loss": 2.5434,
      "step": 379
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.43225640058517456,
      "learning_rate": 0.0009465442699910846,
      "loss": 2.5275,
      "step": 380
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.38182488083839417,
      "learning_rate": 0.0009462094954145801,
      "loss": 2.4222,
      "step": 381
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.34401369094848633,
      "learning_rate": 0.0009458737354018859,
      "loss": 2.5392,
      "step": 382
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3327089548110962,
      "learning_rate": 0.000945536990694515,
      "loss": 2.5488,
      "step": 383
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.38231489062309265,
      "learning_rate": 0.0009451992620361551,
      "loss": 2.5969,
      "step": 384
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.35424381494522095,
      "learning_rate": 0.0009448605501726664,
      "loss": 2.5655,
      "step": 385
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3664548397064209,
      "learning_rate": 0.000944520855852081,
      "loss": 2.4576,
      "step": 386
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.37052857875823975,
      "learning_rate": 0.0009441801798246002,
      "loss": 2.3832,
      "step": 387
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3921554684638977,
      "learning_rate": 0.0009438385228425939,
      "loss": 2.4937,
      "step": 388
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3572741448879242,
      "learning_rate": 0.0009434958856605982,
      "loss": 2.5955,
      "step": 389
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.35618290305137634,
      "learning_rate": 0.0009431522690353137,
      "loss": 2.4845,
      "step": 390
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.35597944259643555,
      "learning_rate": 0.0009428076737256044,
      "loss": 2.4851,
      "step": 391
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3809579908847809,
      "learning_rate": 0.0009424621004924954,
      "loss": 2.5766,
      "step": 392
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3642323315143585,
      "learning_rate": 0.0009421155500991719,
      "loss": 2.5749,
      "step": 393
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.35176482796669006,
      "learning_rate": 0.0009417680233109767,
      "loss": 2.4421,
      "step": 394
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3649900257587433,
      "learning_rate": 0.000941419520895409,
      "loss": 2.5719,
      "step": 395
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3674631118774414,
      "learning_rate": 0.0009410700436221229,
      "loss": 2.5357,
      "step": 396
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3271072506904602,
      "learning_rate": 0.0009407195922629252,
      "loss": 2.4143,
      "step": 397
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3404058814048767,
      "learning_rate": 0.000940368167591774,
      "loss": 2.5296,
      "step": 398
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3525659739971161,
      "learning_rate": 0.0009400157703847769,
      "loss": 2.4324,
      "step": 399
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.34919363260269165,
      "learning_rate": 0.0009396624014201895,
      "loss": 2.5856,
      "step": 400
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3677954077720642,
      "learning_rate": 0.000939308061478413,
      "loss": 2.617,
      "step": 401
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.34744521975517273,
      "learning_rate": 0.0009389527513419935,
      "loss": 2.4802,
      "step": 402
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.35623639822006226,
      "learning_rate": 0.0009385964717956195,
      "loss": 2.5032,
      "step": 403
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.36055660247802734,
      "learning_rate": 0.0009382392236261201,
      "loss": 2.4546,
      "step": 404
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3571677505970001,
      "learning_rate": 0.0009378810076224644,
      "loss": 2.5574,
      "step": 405
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.34384065866470337,
      "learning_rate": 0.0009375218245757582,
      "loss": 2.5043,
      "step": 406
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3361946940422058,
      "learning_rate": 0.0009371616752792432,
      "loss": 2.4553,
      "step": 407
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3476642072200775,
      "learning_rate": 0.0009368005605282949,
      "loss": 2.5216,
      "step": 408
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.345378577709198,
      "learning_rate": 0.0009364384811204212,
      "loss": 2.5408,
      "step": 409
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3399485647678375,
      "learning_rate": 0.00093607543785526,
      "loss": 2.4858,
      "step": 410
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.35123303532600403,
      "learning_rate": 0.0009357114315345787,
      "loss": 2.533,
      "step": 411
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3436707854270935,
      "learning_rate": 0.0009353464629622705,
      "loss": 2.4543,
      "step": 412
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3699854612350464,
      "learning_rate": 0.0009349805329443544,
      "loss": 2.5121,
      "step": 413
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3455789089202881,
      "learning_rate": 0.0009346136422889724,
      "loss": 2.4956,
      "step": 414
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3508444130420685,
      "learning_rate": 0.0009342457918063882,
      "loss": 2.4498,
      "step": 415
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.34029752016067505,
      "learning_rate": 0.0009338769823089853,
      "loss": 2.4936,
      "step": 416
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3479560613632202,
      "learning_rate": 0.0009335072146112648,
      "loss": 2.6382,
      "step": 417
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.357577383518219,
      "learning_rate": 0.0009331364895298444,
      "loss": 2.3616,
      "step": 418
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3635842204093933,
      "learning_rate": 0.0009327648078834559,
      "loss": 2.5604,
      "step": 419
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3546361029148102,
      "learning_rate": 0.0009323921704929434,
      "loss": 2.5032,
      "step": 420
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.36664658784866333,
      "learning_rate": 0.0009320185781812623,
      "loss": 2.3724,
      "step": 421
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3267354369163513,
      "learning_rate": 0.0009316440317734762,
      "loss": 2.4242,
      "step": 422
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3363821804523468,
      "learning_rate": 0.0009312685320967565,
      "loss": 2.546,
      "step": 423
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.34388723969459534,
      "learning_rate": 0.0009308920799803793,
      "loss": 2.3859,
      "step": 424
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.34384873509407043,
      "learning_rate": 0.0009305146762557246,
      "loss": 2.5229,
      "step": 425
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3315981924533844,
      "learning_rate": 0.0009301363217562736,
      "loss": 2.4896,
      "step": 426
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3570757210254669,
      "learning_rate": 0.0009297570173176074,
      "loss": 2.4805,
      "step": 427
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.33402135968208313,
      "learning_rate": 0.000929376763777405,
      "loss": 2.5637,
      "step": 428
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3391449451446533,
      "learning_rate": 0.0009289955619754413,
      "loss": 2.4826,
      "step": 429
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.35202673077583313,
      "learning_rate": 0.0009286134127535859,
      "loss": 2.5871,
      "step": 430
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3385024666786194,
      "learning_rate": 0.0009282303169558,
      "loss": 2.5477,
      "step": 431
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.33344566822052,
      "learning_rate": 0.0009278462754281359,
      "loss": 2.4963,
      "step": 432
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.339773029088974,
      "learning_rate": 0.0009274612890187342,
      "loss": 2.5111,
      "step": 433
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.34970688819885254,
      "learning_rate": 0.0009270753585778222,
      "loss": 2.5259,
      "step": 434
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3609929382801056,
      "learning_rate": 0.0009266884849577124,
      "loss": 2.4784,
      "step": 435
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.34984079003334045,
      "learning_rate": 0.0009263006690127998,
      "loss": 2.4809,
      "step": 436
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.35910046100616455,
      "learning_rate": 0.0009259119115995609,
      "loss": 2.4745,
      "step": 437
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3233133852481842,
      "learning_rate": 0.0009255222135765511,
      "loss": 2.4454,
      "step": 438
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.34542346000671387,
      "learning_rate": 0.0009251315758044032,
      "loss": 2.5045,
      "step": 439
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.35569512844085693,
      "learning_rate": 0.0009247399991458255,
      "loss": 2.4369,
      "step": 440
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3344586193561554,
      "learning_rate": 0.0009243474844655996,
      "loss": 2.5526,
      "step": 441
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3380890190601349,
      "learning_rate": 0.0009239540326305791,
      "loss": 2.4491,
      "step": 442
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3516070246696472,
      "learning_rate": 0.0009235596445096864,
      "loss": 2.5167,
      "step": 443
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3395554721355438,
      "learning_rate": 0.0009231643209739127,
      "loss": 2.474,
      "step": 444
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3406504988670349,
      "learning_rate": 0.0009227680628963145,
      "loss": 2.4949,
      "step": 445
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.457101047039032,
      "learning_rate": 0.000922370871152012,
      "loss": 2.4596,
      "step": 446
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.34821411967277527,
      "learning_rate": 0.0009219727466181877,
      "loss": 2.5031,
      "step": 447
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.32984161376953125,
      "learning_rate": 0.0009215736901740841,
      "loss": 2.4923,
      "step": 448
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.34475263953208923,
      "learning_rate": 0.0009211737027010016,
      "loss": 2.3583,
      "step": 449
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3776776194572449,
      "learning_rate": 0.0009207727850822971,
      "loss": 2.3862,
      "step": 450
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.37130600214004517,
      "learning_rate": 0.0009203709382033814,
      "loss": 2.5682,
      "step": 451
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.34677278995513916,
      "learning_rate": 0.0009199681629517173,
      "loss": 2.5189,
      "step": 452
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.35726290941238403,
      "learning_rate": 0.0009195644602168184,
      "loss": 2.4533,
      "step": 453
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.317497581243515,
      "learning_rate": 0.0009191598308902464,
      "loss": 2.4999,
      "step": 454
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3325786888599396,
      "learning_rate": 0.0009187542758656091,
      "loss": 2.415,
      "step": 455
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.36395639181137085,
      "learning_rate": 0.0009183477960385591,
      "loss": 2.3964,
      "step": 456
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.34762322902679443,
      "learning_rate": 0.0009179403923067912,
      "loss": 2.4957,
      "step": 457
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.38645780086517334,
      "learning_rate": 0.0009175320655700406,
      "loss": 2.4983,
      "step": 458
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3338715136051178,
      "learning_rate": 0.0009171228167300805,
      "loss": 2.4943,
      "step": 459
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3607775568962097,
      "learning_rate": 0.0009167126466907215,
      "loss": 2.4135,
      "step": 460
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3563859164714813,
      "learning_rate": 0.0009163015563578074,
      "loss": 2.5806,
      "step": 461
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.35005345940589905,
      "learning_rate": 0.0009158895466392158,
      "loss": 2.3687,
      "step": 462
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3484789729118347,
      "learning_rate": 0.0009154766184448535,
      "loss": 2.4023,
      "step": 463
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3499147295951843,
      "learning_rate": 0.0009150627726866568,
      "loss": 2.4911,
      "step": 464
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3265226185321808,
      "learning_rate": 0.000914648010278587,
      "loss": 2.5004,
      "step": 465
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3438777029514313,
      "learning_rate": 0.0009142323321366315,
      "loss": 2.502,
      "step": 466
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3390726149082184,
      "learning_rate": 0.0009138157391787986,
      "loss": 2.4414,
      "step": 467
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3248733878135681,
      "learning_rate": 0.0009133982323251177,
      "loss": 2.3809,
      "step": 468
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.33463525772094727,
      "learning_rate": 0.0009129798124976365,
      "loss": 2.5564,
      "step": 469
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.33225491642951965,
      "learning_rate": 0.0009125604806204187,
      "loss": 2.4974,
      "step": 470
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3318762481212616,
      "learning_rate": 0.0009121402376195421,
      "loss": 2.4041,
      "step": 471
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3571879267692566,
      "learning_rate": 0.0009117190844230972,
      "loss": 2.3993,
      "step": 472
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3530467450618744,
      "learning_rate": 0.0009112970219611841,
      "loss": 2.4737,
      "step": 473
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3519071042537689,
      "learning_rate": 0.0009108740511659115,
      "loss": 2.5659,
      "step": 474
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3431400656700134,
      "learning_rate": 0.0009104501729713935,
      "loss": 2.442,
      "step": 475
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.35811561346054077,
      "learning_rate": 0.0009100253883137488,
      "loss": 2.5117,
      "step": 476
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.31780701875686646,
      "learning_rate": 0.0009095996981310974,
      "loss": 2.4335,
      "step": 477
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3387174606323242,
      "learning_rate": 0.0009091731033635596,
      "loss": 2.5276,
      "step": 478
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.34891781210899353,
      "learning_rate": 0.0009087456049532529,
      "loss": 2.4268,
      "step": 479
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4315412640571594,
      "learning_rate": 0.0009083172038442914,
      "loss": 2.4314,
      "step": 480
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3428035080432892,
      "learning_rate": 0.0009078879009827817,
      "loss": 2.3851,
      "step": 481
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.32064807415008545,
      "learning_rate": 0.0009074576973168223,
      "loss": 2.4058,
      "step": 482
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3589988946914673,
      "learning_rate": 0.0009070265937965015,
      "loss": 2.5445,
      "step": 483
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.33878812193870544,
      "learning_rate": 0.0009065945913738942,
      "loss": 2.4102,
      "step": 484
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3507571220397949,
      "learning_rate": 0.0009061616910030609,
      "loss": 2.4323,
      "step": 485
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.32399091124534607,
      "learning_rate": 0.0009057278936400453,
      "loss": 2.3854,
      "step": 486
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3469884693622589,
      "learning_rate": 0.0009052932002428715,
      "loss": 2.641,
      "step": 487
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3163735568523407,
      "learning_rate": 0.0009048576117715435,
      "loss": 2.4769,
      "step": 488
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.33603984117507935,
      "learning_rate": 0.0009044211291880407,
      "loss": 2.34,
      "step": 489
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3335112929344177,
      "learning_rate": 0.000903983753456318,
      "loss": 2.3757,
      "step": 490
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.34720578789711,
      "learning_rate": 0.0009035454855423026,
      "loss": 2.4643,
      "step": 491
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3539136052131653,
      "learning_rate": 0.0009031063264138922,
      "loss": 2.523,
      "step": 492
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.33053672313690186,
      "learning_rate": 0.0009026662770409522,
      "loss": 2.4031,
      "step": 493
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3776102662086487,
      "learning_rate": 0.0009022253383953147,
      "loss": 2.5921,
      "step": 494
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.33027383685112,
      "learning_rate": 0.0009017835114507753,
      "loss": 2.4507,
      "step": 495
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.33334264159202576,
      "learning_rate": 0.0009013407971830914,
      "loss": 2.3747,
      "step": 496
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3344537019729614,
      "learning_rate": 0.0009008971965699801,
      "loss": 2.4686,
      "step": 497
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.33218076825141907,
      "learning_rate": 0.0009004527105911163,
      "loss": 2.4323,
      "step": 498
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3279860019683838,
      "learning_rate": 0.0009000073402281295,
      "loss": 2.4425,
      "step": 499
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.34024667739868164,
      "learning_rate": 0.0008995610864646028,
      "loss": 2.4905,
      "step": 500
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3298107385635376,
      "learning_rate": 0.0008991139502860703,
      "loss": 2.3982,
      "step": 501
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.34255123138427734,
      "learning_rate": 0.0008986659326800146,
      "loss": 2.4431,
      "step": 502
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.32135897874832153,
      "learning_rate": 0.0008982170346358651,
      "loss": 2.4958,
      "step": 503
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.36143195629119873,
      "learning_rate": 0.0008977672571449956,
      "loss": 2.3849,
      "step": 504
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3442467451095581,
      "learning_rate": 0.0008973166012007217,
      "loss": 2.4776,
      "step": 505
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3941558003425598,
      "learning_rate": 0.0008968650677982998,
      "loss": 2.488,
      "step": 506
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.36443281173706055,
      "learning_rate": 0.0008964126579349236,
      "loss": 2.3926,
      "step": 507
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.34229421615600586,
      "learning_rate": 0.0008959593726097226,
      "loss": 2.5584,
      "step": 508
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.324378103017807,
      "learning_rate": 0.0008955052128237596,
      "loss": 2.4542,
      "step": 509
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.33347976207733154,
      "learning_rate": 0.0008950501795800288,
      "loss": 2.4631,
      "step": 510
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.32004567980766296,
      "learning_rate": 0.0008945942738834532,
      "loss": 2.4317,
      "step": 511
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3289385437965393,
      "learning_rate": 0.0008941374967408826,
      "loss": 2.4545,
      "step": 512
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3377652168273926,
      "learning_rate": 0.0008936798491610916,
      "loss": 2.4209,
      "step": 513
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3369903862476349,
      "learning_rate": 0.0008932213321547768,
      "loss": 2.4968,
      "step": 514
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.33552008867263794,
      "learning_rate": 0.0008927619467345554,
      "loss": 2.5045,
      "step": 515
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3757782280445099,
      "learning_rate": 0.0008923016939149615,
      "loss": 2.323,
      "step": 516
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.35516154766082764,
      "learning_rate": 0.0008918405747124458,
      "loss": 2.492,
      "step": 517
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.36546432971954346,
      "learning_rate": 0.0008913785901453721,
      "loss": 2.4578,
      "step": 518
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.34146425127983093,
      "learning_rate": 0.000890915741234015,
      "loss": 2.594,
      "step": 519
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.34310123324394226,
      "learning_rate": 0.0008904520290005582,
      "loss": 2.5402,
      "step": 520
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3336651623249054,
      "learning_rate": 0.000889987454469092,
      "loss": 2.3994,
      "step": 521
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.34388723969459534,
      "learning_rate": 0.0008895220186656111,
      "loss": 2.5464,
      "step": 522
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.33847668766975403,
      "learning_rate": 0.0008890557226180122,
      "loss": 2.4924,
      "step": 523
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.32754337787628174,
      "learning_rate": 0.0008885885673560921,
      "loss": 2.3328,
      "step": 524
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3500807583332062,
      "learning_rate": 0.0008881205539115444,
      "loss": 2.4653,
      "step": 525
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.33577507734298706,
      "learning_rate": 0.0008876516833179589,
      "loss": 2.4518,
      "step": 526
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4069581627845764,
      "learning_rate": 0.0008871819566108177,
      "loss": 2.4127,
      "step": 527
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3372245728969574,
      "learning_rate": 0.000886711374827494,
      "loss": 2.4746,
      "step": 528
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.330152302980423,
      "learning_rate": 0.0008862399390072491,
      "loss": 2.5513,
      "step": 529
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.31766873598098755,
      "learning_rate": 0.0008857676501912305,
      "loss": 2.4002,
      "step": 530
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3010423183441162,
      "learning_rate": 0.0008852945094224697,
      "loss": 2.4351,
      "step": 531
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.31884583830833435,
      "learning_rate": 0.0008848205177458795,
      "loss": 2.4572,
      "step": 532
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.32908159494400024,
      "learning_rate": 0.0008843456762082518,
      "loss": 2.4349,
      "step": 533
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3140951991081238,
      "learning_rate": 0.0008838699858582557,
      "loss": 2.3684,
      "step": 534
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3392878770828247,
      "learning_rate": 0.0008833934477464347,
      "loss": 2.4954,
      "step": 535
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3422383666038513,
      "learning_rate": 0.0008829160629252045,
      "loss": 2.483,
      "step": 536
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3372221291065216,
      "learning_rate": 0.0008824378324488509,
      "loss": 2.4472,
      "step": 537
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.37736964225769043,
      "learning_rate": 0.0008819587573735268,
      "loss": 2.4642,
      "step": 538
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.33233681321144104,
      "learning_rate": 0.0008814788387572513,
      "loss": 2.429,
      "step": 539
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3406238853931427,
      "learning_rate": 0.0008809980776599053,
      "loss": 2.498,
      "step": 540
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.334006667137146,
      "learning_rate": 0.0008805164751432312,
      "loss": 2.4582,
      "step": 541
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.32668536901474,
      "learning_rate": 0.0008800340322708292,
      "loss": 2.4297,
      "step": 542
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.32854482531547546,
      "learning_rate": 0.0008795507501081555,
      "loss": 2.3655,
      "step": 543
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3438929319381714,
      "learning_rate": 0.0008790666297225196,
      "loss": 2.429,
      "step": 544
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.32525596022605896,
      "learning_rate": 0.0008785816721830829,
      "loss": 2.3412,
      "step": 545
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.33793216943740845,
      "learning_rate": 0.0008780958785608546,
      "loss": 2.4113,
      "step": 546
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3195613920688629,
      "learning_rate": 0.0008776092499286912,
      "loss": 2.4335,
      "step": 547
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.32704782485961914,
      "learning_rate": 0.0008771217873612929,
      "loss": 2.335,
      "step": 548
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3418804407119751,
      "learning_rate": 0.0008766334919352017,
      "loss": 2.456,
      "step": 549
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.30939391255378723,
      "learning_rate": 0.0008761443647287987,
      "loss": 2.4692,
      "step": 550
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3286648690700531,
      "learning_rate": 0.0008756544068223026,
      "loss": 2.516,
      "step": 551
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3422103524208069,
      "learning_rate": 0.0008751636192977659,
      "loss": 2.5331,
      "step": 552
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3149074912071228,
      "learning_rate": 0.0008746720032390737,
      "loss": 2.3879,
      "step": 553
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3141325116157532,
      "learning_rate": 0.0008741795597319408,
      "loss": 2.3947,
      "step": 554
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.33686262369155884,
      "learning_rate": 0.0008736862898639095,
      "loss": 2.4111,
      "step": 555
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3246064782142639,
      "learning_rate": 0.0008731921947243468,
      "loss": 2.5171,
      "step": 556
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.34724074602127075,
      "learning_rate": 0.0008726972754044427,
      "loss": 2.4652,
      "step": 557
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.367062509059906,
      "learning_rate": 0.0008722015329972069,
      "loss": 2.4822,
      "step": 558
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3318836987018585,
      "learning_rate": 0.0008717049685974672,
      "loss": 2.502,
      "step": 559
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3711715340614319,
      "learning_rate": 0.0008712075833018665,
      "loss": 2.5164,
      "step": 560
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.31690675020217896,
      "learning_rate": 0.0008707093782088608,
      "loss": 2.3691,
      "step": 561
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.35810965299606323,
      "learning_rate": 0.0008702103544187167,
      "loss": 2.4395,
      "step": 562
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3273511230945587,
      "learning_rate": 0.0008697105130335085,
      "loss": 2.4948,
      "step": 563
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3189830183982849,
      "learning_rate": 0.0008692098551571164,
      "loss": 2.4911,
      "step": 564
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.32062047719955444,
      "learning_rate": 0.0008687083818952235,
      "loss": 2.4449,
      "step": 565
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3242794871330261,
      "learning_rate": 0.0008682060943553143,
      "loss": 2.4272,
      "step": 566
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3219988942146301,
      "learning_rate": 0.0008677029936466707,
      "loss": 2.451,
      "step": 567
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3539111316204071,
      "learning_rate": 0.0008671990808803711,
      "loss": 2.6144,
      "step": 568
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3351738750934601,
      "learning_rate": 0.0008666943571692871,
      "loss": 2.476,
      "step": 569
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.329746276140213,
      "learning_rate": 0.0008661888236280813,
      "loss": 2.4815,
      "step": 570
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.34160110354423523,
      "learning_rate": 0.0008656824813732045,
      "loss": 2.4827,
      "step": 571
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.31913745403289795,
      "learning_rate": 0.000865175331522894,
      "loss": 2.4513,
      "step": 572
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.314571350812912,
      "learning_rate": 0.0008646673751971703,
      "loss": 2.4751,
      "step": 573
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.33775636553764343,
      "learning_rate": 0.000864158613517835,
      "loss": 2.3509,
      "step": 574
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3540123403072357,
      "learning_rate": 0.0008636490476084681,
      "loss": 2.4942,
      "step": 575
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3346134126186371,
      "learning_rate": 0.0008631386785944264,
      "loss": 2.4532,
      "step": 576
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.30813705921173096,
      "learning_rate": 0.0008626275076028397,
      "loss": 2.4586,
      "step": 577
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3027915060520172,
      "learning_rate": 0.0008621155357626091,
      "loss": 2.4227,
      "step": 578
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.32021448016166687,
      "learning_rate": 0.0008616027642044042,
      "loss": 2.4367,
      "step": 579
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.33407455682754517,
      "learning_rate": 0.000861089194060661,
      "loss": 2.3562,
      "step": 580
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3295171856880188,
      "learning_rate": 0.000860574826465579,
      "loss": 2.4507,
      "step": 581
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3145063817501068,
      "learning_rate": 0.0008600596625551191,
      "loss": 2.5406,
      "step": 582
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3258068263530731,
      "learning_rate": 0.0008595437034670006,
      "loss": 2.4065,
      "step": 583
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.32188528776168823,
      "learning_rate": 0.0008590269503406985,
      "loss": 2.4056,
      "step": 584
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.32393890619277954,
      "learning_rate": 0.0008585094043174423,
      "loss": 2.4111,
      "step": 585
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.33656853437423706,
      "learning_rate": 0.0008579910665402118,
      "loss": 2.4731,
      "step": 586
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.33861279487609863,
      "learning_rate": 0.000857471938153736,
      "loss": 2.4013,
      "step": 587
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3394540250301361,
      "learning_rate": 0.0008569520203044892,
      "loss": 2.3817,
      "step": 588
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3245292901992798,
      "learning_rate": 0.0008564313141406901,
      "loss": 2.3469,
      "step": 589
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3375159204006195,
      "learning_rate": 0.0008559098208122973,
      "loss": 2.3767,
      "step": 590
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3351301848888397,
      "learning_rate": 0.0008553875414710089,
      "loss": 2.4174,
      "step": 591
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3307935297489166,
      "learning_rate": 0.0008548644772702579,
      "loss": 2.3402,
      "step": 592
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.31705746054649353,
      "learning_rate": 0.0008543406293652116,
      "loss": 2.4362,
      "step": 593
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.31638142466545105,
      "learning_rate": 0.0008538159989127671,
      "loss": 2.4184,
      "step": 594
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.32090166211128235,
      "learning_rate": 0.0008532905870715505,
      "loss": 2.4351,
      "step": 595
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3248273730278015,
      "learning_rate": 0.0008527643950019131,
      "loss": 2.3781,
      "step": 596
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3112647831439972,
      "learning_rate": 0.0008522374238659296,
      "loss": 2.4181,
      "step": 597
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.32619887590408325,
      "learning_rate": 0.0008517096748273951,
      "loss": 2.4827,
      "step": 598
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3540405333042145,
      "learning_rate": 0.0008511811490518227,
      "loss": 2.4757,
      "step": 599
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.32304564118385315,
      "learning_rate": 0.0008506518477064405,
      "loss": 2.4868,
      "step": 600
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3260352909564972,
      "learning_rate": 0.0008501217719601903,
      "loss": 2.4632,
      "step": 601
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3379616141319275,
      "learning_rate": 0.0008495909229837233,
      "loss": 2.4834,
      "step": 602
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3170284628868103,
      "learning_rate": 0.000849059301949399,
      "loss": 2.3621,
      "step": 603
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.32728955149650574,
      "learning_rate": 0.0008485269100312812,
      "loss": 2.3993,
      "step": 604
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.32943764328956604,
      "learning_rate": 0.0008479937484051368,
      "loss": 2.4752,
      "step": 605
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3163626492023468,
      "learning_rate": 0.0008474598182484323,
      "loss": 2.3921,
      "step": 606
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.32608652114868164,
      "learning_rate": 0.0008469251207403317,
      "loss": 2.4796,
      "step": 607
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.361741840839386,
      "learning_rate": 0.0008463896570616934,
      "loss": 2.478,
      "step": 608
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.32413479685783386,
      "learning_rate": 0.0008458534283950678,
      "loss": 2.4267,
      "step": 609
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3459324538707733,
      "learning_rate": 0.0008453164359246952,
      "loss": 2.5862,
      "step": 610
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3310210704803467,
      "learning_rate": 0.0008447786808365022,
      "loss": 2.4723,
      "step": 611
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.30966681241989136,
      "learning_rate": 0.0008442401643181,
      "loss": 2.4603,
      "step": 612
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.30379775166511536,
      "learning_rate": 0.0008437008875587811,
      "loss": 2.4703,
      "step": 613
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.31512007117271423,
      "learning_rate": 0.0008431608517495171,
      "loss": 2.4754,
      "step": 614
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3118310868740082,
      "learning_rate": 0.0008426200580829561,
      "loss": 2.3377,
      "step": 615
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3341467082500458,
      "learning_rate": 0.0008420785077534195,
      "loss": 2.5572,
      "step": 616
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3005862534046173,
      "learning_rate": 0.0008415362019569001,
      "loss": 2.3373,
      "step": 617
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.30986499786376953,
      "learning_rate": 0.0008409931418910591,
      "loss": 2.4026,
      "step": 618
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.33998921513557434,
      "learning_rate": 0.0008404493287552232,
      "loss": 2.4342,
      "step": 619
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3119550943374634,
      "learning_rate": 0.0008399047637503825,
      "loss": 2.4336,
      "step": 620
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.31371214985847473,
      "learning_rate": 0.0008393594480791875,
      "loss": 2.4541,
      "step": 621
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3188551366329193,
      "learning_rate": 0.0008388133829459463,
      "loss": 2.4322,
      "step": 622
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3495667576789856,
      "learning_rate": 0.0008382665695566227,
      "loss": 2.4277,
      "step": 623
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3576057553291321,
      "learning_rate": 0.0008377190091188324,
      "loss": 2.4557,
      "step": 624
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.32205551862716675,
      "learning_rate": 0.0008371707028418413,
      "loss": 2.4358,
      "step": 625
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.33084583282470703,
      "learning_rate": 0.0008366216519365621,
      "loss": 2.4613,
      "step": 626
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.30437248945236206,
      "learning_rate": 0.0008360718576155525,
      "loss": 2.4816,
      "step": 627
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3244648575782776,
      "learning_rate": 0.0008355213210930118,
      "loss": 2.4144,
      "step": 628
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3115200996398926,
      "learning_rate": 0.0008349700435847778,
      "loss": 2.4452,
      "step": 629
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3367325961589813,
      "learning_rate": 0.0008344180263083256,
      "loss": 2.5294,
      "step": 630
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3251790404319763,
      "learning_rate": 0.000833865270482764,
      "loss": 2.4928,
      "step": 631
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.4025709331035614,
      "learning_rate": 0.0008333117773288324,
      "loss": 2.4067,
      "step": 632
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3342664837837219,
      "learning_rate": 0.0008327575480688985,
      "loss": 2.4376,
      "step": 633
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3202410042285919,
      "learning_rate": 0.000832202583926956,
      "loss": 2.4857,
      "step": 634
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3315035104751587,
      "learning_rate": 0.0008316468861286217,
      "loss": 2.4438,
      "step": 635
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3262021541595459,
      "learning_rate": 0.0008310904559011323,
      "loss": 2.4837,
      "step": 636
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3089523911476135,
      "learning_rate": 0.0008305332944733419,
      "loss": 2.4862,
      "step": 637
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.31825029850006104,
      "learning_rate": 0.0008299754030757202,
      "loss": 2.4015,
      "step": 638
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.31092581152915955,
      "learning_rate": 0.0008294167829403481,
      "loss": 2.3761,
      "step": 639
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.32116973400115967,
      "learning_rate": 0.0008288574353009164,
      "loss": 2.4585,
      "step": 640
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3136792480945587,
      "learning_rate": 0.0008282973613927225,
      "loss": 2.3248,
      "step": 641
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2995297908782959,
      "learning_rate": 0.0008277365624526675,
      "loss": 2.2899,
      "step": 642
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3079420328140259,
      "learning_rate": 0.0008271750397192541,
      "loss": 2.4246,
      "step": 643
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.30741819739341736,
      "learning_rate": 0.0008266127944325832,
      "loss": 2.2536,
      "step": 644
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3283785879611969,
      "learning_rate": 0.0008260498278343513,
      "loss": 2.4228,
      "step": 645
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3278549313545227,
      "learning_rate": 0.0008254861411678485,
      "loss": 2.4684,
      "step": 646
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.33177945017814636,
      "learning_rate": 0.0008249217356779544,
      "loss": 2.4754,
      "step": 647
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3355714976787567,
      "learning_rate": 0.0008243566126111363,
      "loss": 2.3887,
      "step": 648
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.31164517998695374,
      "learning_rate": 0.0008237907732154466,
      "loss": 2.4165,
      "step": 649
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3204694092273712,
      "learning_rate": 0.0008232242187405194,
      "loss": 2.4273,
      "step": 650
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.31102320551872253,
      "learning_rate": 0.000822656950437568,
      "loss": 2.4354,
      "step": 651
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3352205157279968,
      "learning_rate": 0.0008220889695593823,
      "loss": 2.39,
      "step": 652
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3436696529388428,
      "learning_rate": 0.0008215202773603259,
      "loss": 2.4163,
      "step": 653
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3375329375267029,
      "learning_rate": 0.0008209508750963328,
      "loss": 2.3571,
      "step": 654
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.34684598445892334,
      "learning_rate": 0.0008203807640249062,
      "loss": 2.44,
      "step": 655
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3398624360561371,
      "learning_rate": 0.0008198099454051136,
      "loss": 2.4729,
      "step": 656
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3295707106590271,
      "learning_rate": 0.0008192384204975857,
      "loss": 2.4132,
      "step": 657
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3300459086894989,
      "learning_rate": 0.000818666190564513,
      "loss": 2.403,
      "step": 658
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3090726435184479,
      "learning_rate": 0.0008180932568696426,
      "loss": 2.3316,
      "step": 659
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3058747351169586,
      "learning_rate": 0.0008175196206782764,
      "loss": 2.3259,
      "step": 660
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.32757681608200073,
      "learning_rate": 0.0008169452832572675,
      "loss": 2.4858,
      "step": 661
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.33550572395324707,
      "learning_rate": 0.0008163702458750173,
      "loss": 2.5296,
      "step": 662
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3205187916755676,
      "learning_rate": 0.0008157945098014734,
      "loss": 2.3391,
      "step": 663
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.30833372473716736,
      "learning_rate": 0.0008152180763081267,
      "loss": 2.3805,
      "step": 664
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.32461127638816833,
      "learning_rate": 0.0008146409466680076,
      "loss": 2.3504,
      "step": 665
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.317560076713562,
      "learning_rate": 0.0008140631221556845,
      "loss": 2.4497,
      "step": 666
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3331639766693115,
      "learning_rate": 0.0008134846040472599,
      "loss": 2.4971,
      "step": 667
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3365515172481537,
      "learning_rate": 0.0008129053936203688,
      "loss": 2.4447,
      "step": 668
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.33443230390548706,
      "learning_rate": 0.0008123254921541745,
      "loss": 2.4305,
      "step": 669
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.313658744096756,
      "learning_rate": 0.0008117449009293668,
      "loss": 2.2522,
      "step": 670
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.32625067234039307,
      "learning_rate": 0.0008111636212281586,
      "loss": 2.4191,
      "step": 671
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3078235685825348,
      "learning_rate": 0.0008105816543342833,
      "loss": 2.523,
      "step": 672
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.34956037998199463,
      "learning_rate": 0.0008099990015329919,
      "loss": 2.415,
      "step": 673
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3143714964389801,
      "learning_rate": 0.0008094156641110504,
      "loss": 2.3877,
      "step": 674
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3060227334499359,
      "learning_rate": 0.0008088316433567369,
      "loss": 2.4795,
      "step": 675
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3220013380050659,
      "learning_rate": 0.0008082469405598378,
      "loss": 2.4475,
      "step": 676
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3233586549758911,
      "learning_rate": 0.0008076615570116468,
      "loss": 2.4476,
      "step": 677
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3025561273097992,
      "learning_rate": 0.0008070754940049603,
      "loss": 2.3455,
      "step": 678
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.33245453238487244,
      "learning_rate": 0.0008064887528340756,
      "loss": 2.5279,
      "step": 679
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3264337182044983,
      "learning_rate": 0.0008059013347947874,
      "loss": 2.4374,
      "step": 680
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.31236332654953003,
      "learning_rate": 0.0008053132411843857,
      "loss": 2.487,
      "step": 681
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.2997890114784241,
      "learning_rate": 0.0008047244733016521,
      "loss": 2.4395,
      "step": 682
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.32226288318634033,
      "learning_rate": 0.0008041350324468573,
      "loss": 2.357,
      "step": 683
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3150307536125183,
      "learning_rate": 0.0008035449199217583,
      "loss": 2.4903,
      "step": 684
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.31021836400032043,
      "learning_rate": 0.0008029541370295957,
      "loss": 2.3422,
      "step": 685
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.30424338579177856,
      "learning_rate": 0.0008023626850750903,
      "loss": 2.319,
      "step": 686
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3468063771724701,
      "learning_rate": 0.0008017705653644406,
      "loss": 2.5002,
      "step": 687
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3688942790031433,
      "learning_rate": 0.0008011777792053195,
      "loss": 2.4917,
      "step": 688
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.32205426692962646,
      "learning_rate": 0.0008005843279068725,
      "loss": 2.3285,
      "step": 689
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.32176074385643005,
      "learning_rate": 0.000799990212779713,
      "loss": 2.5435,
      "step": 690
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.32653990387916565,
      "learning_rate": 0.0007993954351359214,
      "loss": 2.4945,
      "step": 691
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.2893826961517334,
      "learning_rate": 0.0007987999962890406,
      "loss": 2.381,
      "step": 692
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.31107470393180847,
      "learning_rate": 0.0007982038975540742,
      "loss": 2.4155,
      "step": 693
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.41854846477508545,
      "learning_rate": 0.0007976071402474826,
      "loss": 2.5316,
      "step": 694
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.31840115785598755,
      "learning_rate": 0.0007970097256871811,
      "loss": 2.4409,
      "step": 695
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.323106974363327,
      "learning_rate": 0.0007964116551925364,
      "loss": 2.4463,
      "step": 696
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.32400813698768616,
      "learning_rate": 0.0007958129300843637,
      "loss": 2.3519,
      "step": 697
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.30560702085494995,
      "learning_rate": 0.0007952135516849239,
      "loss": 2.3795,
      "step": 698
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.33904680609703064,
      "learning_rate": 0.0007946135213179207,
      "loss": 2.4355,
      "step": 699
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.34928417205810547,
      "learning_rate": 0.0007940128403084977,
      "loss": 2.4738,
      "step": 700
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.34188196063041687,
      "learning_rate": 0.0007934115099832355,
      "loss": 2.4211,
      "step": 701
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.30917805433273315,
      "learning_rate": 0.0007928095316701483,
      "loss": 2.4305,
      "step": 702
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.33448168635368347,
      "learning_rate": 0.0007922069066986819,
      "loss": 2.3273,
      "step": 703
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3173739016056061,
      "learning_rate": 0.0007916036363997097,
      "loss": 2.4743,
      "step": 704
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3055970072746277,
      "learning_rate": 0.0007909997221055308,
      "loss": 2.2865,
      "step": 705
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.32079920172691345,
      "learning_rate": 0.0007903951651498658,
      "loss": 2.3032,
      "step": 706
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.31803372502326965,
      "learning_rate": 0.0007897899668678557,
      "loss": 2.4368,
      "step": 707
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.31907787919044495,
      "learning_rate": 0.0007891841285960566,
      "loss": 2.4122,
      "step": 708
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3257612884044647,
      "learning_rate": 0.0007885776516724388,
      "loss": 2.4267,
      "step": 709
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.29811665415763855,
      "learning_rate": 0.0007879705374363831,
      "loss": 2.5159,
      "step": 710
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3139455318450928,
      "learning_rate": 0.000787362787228677,
      "loss": 2.4195,
      "step": 711
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.29211291670799255,
      "learning_rate": 0.0007867544023915134,
      "loss": 2.3421,
      "step": 712
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.32226380705833435,
      "learning_rate": 0.0007861453842684861,
      "loss": 2.3821,
      "step": 713
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3239883482456207,
      "learning_rate": 0.0007855357342045882,
      "loss": 2.3008,
      "step": 714
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.33186987042427063,
      "learning_rate": 0.0007849254535462074,
      "loss": 2.5408,
      "step": 715
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3391883075237274,
      "learning_rate": 0.0007843145436411252,
      "loss": 2.4149,
      "step": 716
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3518095910549164,
      "learning_rate": 0.0007837030058385117,
      "loss": 2.3567,
      "step": 717
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.331024706363678,
      "learning_rate": 0.0007830908414889246,
      "loss": 2.4259,
      "step": 718
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3159828782081604,
      "learning_rate": 0.0007824780519443046,
      "loss": 2.4745,
      "step": 719
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.31883201003074646,
      "learning_rate": 0.0007818646385579735,
      "loss": 2.4183,
      "step": 720
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.31947025656700134,
      "learning_rate": 0.0007812506026846307,
      "loss": 2.3119,
      "step": 721
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3259248733520508,
      "learning_rate": 0.0007806359456803504,
      "loss": 2.4509,
      "step": 722
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3436729907989502,
      "learning_rate": 0.0007800206689025785,
      "loss": 2.4423,
      "step": 723
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3946915864944458,
      "learning_rate": 0.0007794047737101297,
      "loss": 2.3791,
      "step": 724
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.362779438495636,
      "learning_rate": 0.0007787882614631843,
      "loss": 2.3762,
      "step": 725
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.32816633582115173,
      "learning_rate": 0.0007781711335232856,
      "loss": 2.4029,
      "step": 726
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3259948790073395,
      "learning_rate": 0.0007775533912533363,
      "loss": 2.4172,
      "step": 727
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3020377457141876,
      "learning_rate": 0.0007769350360175962,
      "loss": 2.4752,
      "step": 728
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3697303533554077,
      "learning_rate": 0.0007763160691816784,
      "loss": 2.4582,
      "step": 729
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3364540636539459,
      "learning_rate": 0.000775696492112547,
      "loss": 2.4027,
      "step": 730
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.33001646399497986,
      "learning_rate": 0.0007750763061785137,
      "loss": 2.4846,
      "step": 731
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3214930295944214,
      "learning_rate": 0.000774455512749235,
      "loss": 2.2982,
      "step": 732
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3295876383781433,
      "learning_rate": 0.0007738341131957085,
      "loss": 2.4099,
      "step": 733
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3436155915260315,
      "learning_rate": 0.000773212108890271,
      "loss": 2.3752,
      "step": 734
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.33042678236961365,
      "learning_rate": 0.0007725895012065947,
      "loss": 2.4857,
      "step": 735
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3182051479816437,
      "learning_rate": 0.0007719662915196844,
      "loss": 2.331,
      "step": 736
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.31312447786331177,
      "learning_rate": 0.0007713424812058736,
      "loss": 2.3748,
      "step": 737
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.31449154019355774,
      "learning_rate": 0.0007707180716428237,
      "loss": 2.3708,
      "step": 738
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.31184959411621094,
      "learning_rate": 0.0007700930642095184,
      "loss": 2.3877,
      "step": 739
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.32068321108818054,
      "learning_rate": 0.0007694674602862621,
      "loss": 2.4353,
      "step": 740
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.32020846009254456,
      "learning_rate": 0.0007688412612546769,
      "loss": 2.5051,
      "step": 741
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3182331919670105,
      "learning_rate": 0.0007682144684976983,
      "loss": 2.4119,
      "step": 742
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3066627085208893,
      "learning_rate": 0.0007675870833995739,
      "loss": 2.4607,
      "step": 743
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3132324516773224,
      "learning_rate": 0.0007669591073458592,
      "loss": 2.4694,
      "step": 744
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3069417178630829,
      "learning_rate": 0.0007663305417234146,
      "loss": 2.474,
      "step": 745
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3084588050842285,
      "learning_rate": 0.0007657013879204022,
      "loss": 2.3888,
      "step": 746
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3260480761528015,
      "learning_rate": 0.0007650716473262842,
      "loss": 2.3181,
      "step": 747
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3289788067340851,
      "learning_rate": 0.0007644413213318177,
      "loss": 2.431,
      "step": 748
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.33380210399627686,
      "learning_rate": 0.0007638104113290531,
      "loss": 2.4,
      "step": 749
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3170608878135681,
      "learning_rate": 0.0007631789187113303,
      "loss": 2.3853,
      "step": 750
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.31449198722839355,
      "learning_rate": 0.000762546844873276,
      "loss": 2.3422,
      "step": 751
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3366459906101227,
      "learning_rate": 0.0007619141912108007,
      "loss": 2.5217,
      "step": 752
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.32196545600891113,
      "learning_rate": 0.000761280959121095,
      "loss": 2.3469,
      "step": 753
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.30045849084854126,
      "learning_rate": 0.0007606471500026273,
      "loss": 2.4233,
      "step": 754
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3027009963989258,
      "learning_rate": 0.0007600127652551401,
      "loss": 2.4435,
      "step": 755
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.32681769132614136,
      "learning_rate": 0.0007593778062796472,
      "loss": 2.392,
      "step": 756
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.29754146933555603,
      "learning_rate": 0.000758742274478431,
      "loss": 2.3776,
      "step": 757
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.32818469405174255,
      "learning_rate": 0.0007581061712550381,
      "loss": 2.5188,
      "step": 758
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3157346844673157,
      "learning_rate": 0.0007574694980142779,
      "loss": 2.4173,
      "step": 759
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3182338774204254,
      "learning_rate": 0.0007568322561622183,
      "loss": 2.4432,
      "step": 760
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.312049001455307,
      "learning_rate": 0.0007561944471061826,
      "loss": 2.391,
      "step": 761
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.31123027205467224,
      "learning_rate": 0.0007555560722547475,
      "loss": 2.4623,
      "step": 762
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3220462203025818,
      "learning_rate": 0.0007549171330177387,
      "loss": 2.3381,
      "step": 763
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.29605549573898315,
      "learning_rate": 0.0007542776308062285,
      "loss": 2.4709,
      "step": 764
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3164387047290802,
      "learning_rate": 0.0007536375670325325,
      "loss": 2.4393,
      "step": 765
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3019196689128876,
      "learning_rate": 0.0007529969431102063,
      "loss": 2.3663,
      "step": 766
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.31063133478164673,
      "learning_rate": 0.000752355760454043,
      "loss": 2.3799,
      "step": 767
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3407577574253082,
      "learning_rate": 0.0007517140204800693,
      "loss": 2.4195,
      "step": 768
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3127473294734955,
      "learning_rate": 0.0007510717246055425,
      "loss": 2.4036,
      "step": 769
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.33271944522857666,
      "learning_rate": 0.0007504288742489482,
      "loss": 2.38,
      "step": 770
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.31853505969047546,
      "learning_rate": 0.0007497854708299963,
      "loss": 2.4996,
      "step": 771
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.2955154478549957,
      "learning_rate": 0.0007491415157696178,
      "loss": 2.4004,
      "step": 772
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.33185699582099915,
      "learning_rate": 0.0007484970104899623,
      "loss": 2.3859,
      "step": 773
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.2997734844684601,
      "learning_rate": 0.0007478519564143945,
      "loss": 2.5396,
      "step": 774
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3060145378112793,
      "learning_rate": 0.000747206354967491,
      "loss": 2.4496,
      "step": 775
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.30696845054626465,
      "learning_rate": 0.0007465602075750373,
      "loss": 2.5204,
      "step": 776
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3670702576637268,
      "learning_rate": 0.0007459135156640247,
      "loss": 2.3723,
      "step": 777
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3117799162864685,
      "learning_rate": 0.0007452662806626468,
      "loss": 2.3794,
      "step": 778
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.2943999469280243,
      "learning_rate": 0.0007446185040002967,
      "loss": 2.3565,
      "step": 779
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.32253140211105347,
      "learning_rate": 0.0007439701871075642,
      "loss": 2.4588,
      "step": 780
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.32627058029174805,
      "learning_rate": 0.0007433213314162313,
      "loss": 2.3867,
      "step": 781
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3090547025203705,
      "learning_rate": 0.0007426719383592705,
      "loss": 2.4566,
      "step": 782
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.32847535610198975,
      "learning_rate": 0.000742022009370841,
      "loss": 2.3945,
      "step": 783
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.36413154006004333,
      "learning_rate": 0.0007413715458862855,
      "loss": 2.3633,
      "step": 784
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.31347307562828064,
      "learning_rate": 0.0007407205493421272,
      "loss": 2.3775,
      "step": 785
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3196096420288086,
      "learning_rate": 0.0007400690211760661,
      "loss": 2.3772,
      "step": 786
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.29972848296165466,
      "learning_rate": 0.0007394169628269771,
      "loss": 2.3707,
      "step": 787
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.303053617477417,
      "learning_rate": 0.0007387643757349051,
      "loss": 2.4096,
      "step": 788
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.32429173588752747,
      "learning_rate": 0.0007381112613410635,
      "loss": 2.4685,
      "step": 789
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.30890965461730957,
      "learning_rate": 0.0007374576210878298,
      "loss": 2.3896,
      "step": 790
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3229944407939911,
      "learning_rate": 0.0007368034564187425,
      "loss": 2.4195,
      "step": 791
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3196941018104553,
      "learning_rate": 0.0007361487687784989,
      "loss": 2.4215,
      "step": 792
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.32351216673851013,
      "learning_rate": 0.0007354935596129513,
      "loss": 2.4025,
      "step": 793
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.33310917019844055,
      "learning_rate": 0.000734837830369103,
      "loss": 2.4182,
      "step": 794
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3048180341720581,
      "learning_rate": 0.0007341815824951066,
      "loss": 2.3808,
      "step": 795
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3233805000782013,
      "learning_rate": 0.0007335248174402597,
      "loss": 2.4186,
      "step": 796
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3019264042377472,
      "learning_rate": 0.0007328675366550023,
      "loss": 2.4306,
      "step": 797
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3088935613632202,
      "learning_rate": 0.0007322097415909134,
      "loss": 2.3334,
      "step": 798
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3188239634037018,
      "learning_rate": 0.0007315514337007071,
      "loss": 2.4357,
      "step": 799
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3220513164997101,
      "learning_rate": 0.0007308926144382312,
      "loss": 2.3953,
      "step": 800
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.30522820353507996,
      "learning_rate": 0.0007302332852584619,
      "loss": 2.3562,
      "step": 801
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.30965691804885864,
      "learning_rate": 0.0007295734476175018,
      "loss": 2.488,
      "step": 802
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2992212176322937,
      "learning_rate": 0.0007289131029725768,
      "loss": 2.4257,
      "step": 803
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2950397729873657,
      "learning_rate": 0.0007282522527820319,
      "loss": 2.4194,
      "step": 804
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3153960704803467,
      "learning_rate": 0.000727590898505329,
      "loss": 2.3768,
      "step": 805
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.309699147939682,
      "learning_rate": 0.0007269290416030429,
      "loss": 2.2753,
      "step": 806
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.33466726541519165,
      "learning_rate": 0.000726266683536859,
      "loss": 2.3223,
      "step": 807
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3174177408218384,
      "learning_rate": 0.0007256038257695687,
      "loss": 2.3787,
      "step": 808
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.31052225828170776,
      "learning_rate": 0.0007249404697650678,
      "loss": 2.4529,
      "step": 809
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3003176152706146,
      "learning_rate": 0.0007242766169883518,
      "loss": 2.4157,
      "step": 810
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.31503522396087646,
      "learning_rate": 0.0007236122689055138,
      "loss": 2.4934,
      "step": 811
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.310621976852417,
      "learning_rate": 0.0007229474269837401,
      "loss": 2.4226,
      "step": 812
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.30966752767562866,
      "learning_rate": 0.0007222820926913085,
      "loss": 2.3484,
      "step": 813
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3062936067581177,
      "learning_rate": 0.0007216162674975833,
      "loss": 2.3875,
      "step": 814
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.30515366792678833,
      "learning_rate": 0.0007209499528730138,
      "loss": 2.4396,
      "step": 815
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.29733240604400635,
      "learning_rate": 0.0007202831502891294,
      "loss": 2.331,
      "step": 816
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3088294267654419,
      "learning_rate": 0.0007196158612185375,
      "loss": 2.4347,
      "step": 817
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.29479148983955383,
      "learning_rate": 0.0007189480871349201,
      "loss": 2.2831,
      "step": 818
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.314521849155426,
      "learning_rate": 0.0007182798295130299,
      "loss": 2.2739,
      "step": 819
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3179396688938141,
      "learning_rate": 0.0007176110898286878,
      "loss": 2.2614,
      "step": 820
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2973566949367523,
      "learning_rate": 0.0007169418695587791,
      "loss": 2.2874,
      "step": 821
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3212588131427765,
      "learning_rate": 0.0007162721701812506,
      "loss": 2.3673,
      "step": 822
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3252088725566864,
      "learning_rate": 0.0007156019931751072,
      "loss": 2.3185,
      "step": 823
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.29303184151649475,
      "learning_rate": 0.0007149313400204082,
      "loss": 2.3844,
      "step": 824
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.30406293272972107,
      "learning_rate": 0.0007142602121982653,
      "loss": 2.3524,
      "step": 825
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2986964285373688,
      "learning_rate": 0.0007135886111908379,
      "loss": 2.4591,
      "step": 826
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.29020172357559204,
      "learning_rate": 0.0007129165384813303,
      "loss": 2.4148,
      "step": 827
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2968149185180664,
      "learning_rate": 0.0007122439955539888,
      "loss": 2.4637,
      "step": 828
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3008701801300049,
      "learning_rate": 0.0007115709838940983,
      "loss": 2.425,
      "step": 829
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.29755234718322754,
      "learning_rate": 0.0007108975049879785,
      "loss": 2.377,
      "step": 830
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2999955713748932,
      "learning_rate": 0.0007102235603229814,
      "loss": 2.3369,
      "step": 831
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.29546597599983215,
      "learning_rate": 0.000709549151387487,
      "loss": 2.3545,
      "step": 832
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3079194128513336,
      "learning_rate": 0.0007088742796709013,
      "loss": 2.4039,
      "step": 833
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3104572594165802,
      "learning_rate": 0.000708198946663652,
      "loss": 2.3692,
      "step": 834
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3274907171726227,
      "learning_rate": 0.0007075231538571856,
      "loss": 2.4363,
      "step": 835
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3110281825065613,
      "learning_rate": 0.0007068469027439641,
      "loss": 2.5114,
      "step": 836
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.30068644881248474,
      "learning_rate": 0.0007061701948174613,
      "loss": 2.453,
      "step": 837
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3134911060333252,
      "learning_rate": 0.0007054930315721606,
      "loss": 2.3706,
      "step": 838
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3053911626338959,
      "learning_rate": 0.0007048154145035501,
      "loss": 2.402,
      "step": 839
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3110945522785187,
      "learning_rate": 0.0007041373451081207,
      "loss": 2.4252,
      "step": 840
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3051478862762451,
      "learning_rate": 0.0007034588248833621,
      "loss": 2.4945,
      "step": 841
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.30409401655197144,
      "learning_rate": 0.0007027798553277595,
      "loss": 2.3606,
      "step": 842
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.30384305119514465,
      "learning_rate": 0.0007021004379407909,
      "loss": 2.3634,
      "step": 843
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.32497265934944153,
      "learning_rate": 0.0007014205742229227,
      "loss": 2.464,
      "step": 844
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.31966155767440796,
      "learning_rate": 0.0007007402656756072,
      "loss": 2.4033,
      "step": 845
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.31709787249565125,
      "learning_rate": 0.0007000595138012797,
      "loss": 2.4478,
      "step": 846
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.32535532116889954,
      "learning_rate": 0.0006993783201033535,
      "loss": 2.3214,
      "step": 847
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3213086724281311,
      "learning_rate": 0.0006986966860862182,
      "loss": 2.4239,
      "step": 848
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.31199124455451965,
      "learning_rate": 0.000698014613255236,
      "loss": 2.3243,
      "step": 849
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.301448792219162,
      "learning_rate": 0.0006973321031167382,
      "loss": 2.428,
      "step": 850
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3197269141674042,
      "learning_rate": 0.0006966491571780216,
      "loss": 2.4562,
      "step": 851
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3240511119365692,
      "learning_rate": 0.0006959657769473453,
      "loss": 2.3331,
      "step": 852
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.31758731603622437,
      "learning_rate": 0.000695281963933928,
      "loss": 2.378,
      "step": 853
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3192739486694336,
      "learning_rate": 0.0006945977196479438,
      "loss": 2.3175,
      "step": 854
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.30350473523139954,
      "learning_rate": 0.0006939130456005196,
      "loss": 2.3825,
      "step": 855
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.32667604088783264,
      "learning_rate": 0.0006932279433037311,
      "loss": 2.391,
      "step": 856
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3384980857372284,
      "learning_rate": 0.0006925424142705997,
      "loss": 2.3989,
      "step": 857
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2995279133319855,
      "learning_rate": 0.0006918564600150896,
      "loss": 2.3668,
      "step": 858
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.29280999302864075,
      "learning_rate": 0.0006911700820521042,
      "loss": 2.3573,
      "step": 859
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.29242753982543945,
      "learning_rate": 0.0006904832818974818,
      "loss": 2.3121,
      "step": 860
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.29296863079071045,
      "learning_rate": 0.0006897960610679939,
      "loss": 2.3884,
      "step": 861
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.29147234559059143,
      "learning_rate": 0.0006891084210813407,
      "loss": 2.2005,
      "step": 862
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.29946398735046387,
      "learning_rate": 0.0006884203634561483,
      "loss": 2.313,
      "step": 863
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3013591468334198,
      "learning_rate": 0.0006877318897119651,
      "loss": 2.3741,
      "step": 864
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.30276021361351013,
      "learning_rate": 0.0006870430013692579,
      "loss": 2.3945,
      "step": 865
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2938203513622284,
      "learning_rate": 0.0006863536999494101,
      "loss": 2.3457,
      "step": 866
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.30481842160224915,
      "learning_rate": 0.0006856639869747167,
      "loss": 2.3723,
      "step": 867
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.32060059905052185,
      "learning_rate": 0.0006849738639683818,
      "loss": 2.3229,
      "step": 868
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.32711663842201233,
      "learning_rate": 0.000684283332454515,
      "loss": 2.3901,
      "step": 869
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.32909560203552246,
      "learning_rate": 0.0006835923939581281,
      "loss": 2.4458,
      "step": 870
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.310490220785141,
      "learning_rate": 0.0006829010500051318,
      "loss": 2.3041,
      "step": 871
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3057289123535156,
      "learning_rate": 0.0006822093021223321,
      "loss": 2.2987,
      "step": 872
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.32579588890075684,
      "learning_rate": 0.0006815171518374268,
      "loss": 2.3582,
      "step": 873
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.321662575006485,
      "learning_rate": 0.0006808246006790031,
      "loss": 2.4506,
      "step": 874
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.32636263966560364,
      "learning_rate": 0.0006801316501765329,
      "loss": 2.475,
      "step": 875
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3354184627532959,
      "learning_rate": 0.0006794383018603704,
      "loss": 2.3616,
      "step": 876
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3026958405971527,
      "learning_rate": 0.0006787445572617481,
      "loss": 2.4398,
      "step": 877
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.30630719661712646,
      "learning_rate": 0.0006780504179127734,
      "loss": 2.3802,
      "step": 878
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.30914878845214844,
      "learning_rate": 0.0006773558853464265,
      "loss": 2.3894,
      "step": 879
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2927555441856384,
      "learning_rate": 0.000676660961096555,
      "loss": 2.4317,
      "step": 880
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.31402260065078735,
      "learning_rate": 0.000675965646697872,
      "loss": 2.3192,
      "step": 881
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.29512402415275574,
      "learning_rate": 0.0006752699436859519,
      "loss": 2.2709,
      "step": 882
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3162606656551361,
      "learning_rate": 0.0006745738535972279,
      "loss": 2.353,
      "step": 883
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3331770598888397,
      "learning_rate": 0.0006738773779689874,
      "loss": 2.4027,
      "step": 884
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3455560803413391,
      "learning_rate": 0.0006731805183393696,
      "loss": 2.3484,
      "step": 885
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.32219162583351135,
      "learning_rate": 0.0006724832762473618,
      "loss": 2.4197,
      "step": 886
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.31553009152412415,
      "learning_rate": 0.0006717856532327956,
      "loss": 2.3765,
      "step": 887
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3095285892486572,
      "learning_rate": 0.0006710876508363444,
      "loss": 2.3863,
      "step": 888
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3159654140472412,
      "learning_rate": 0.0006703892705995189,
      "loss": 2.2728,
      "step": 889
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3302440643310547,
      "learning_rate": 0.0006696905140646647,
      "loss": 2.4535,
      "step": 890
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.304210901260376,
      "learning_rate": 0.0006689913827749581,
      "loss": 2.4112,
      "step": 891
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.30010223388671875,
      "learning_rate": 0.0006682918782744032,
      "loss": 2.4155,
      "step": 892
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3095526397228241,
      "learning_rate": 0.0006675920021078282,
      "loss": 2.415,
      "step": 893
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3131411373615265,
      "learning_rate": 0.0006668917558208823,
      "loss": 2.4055,
      "step": 894
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.30869826674461365,
      "learning_rate": 0.0006661911409600321,
      "loss": 2.4512,
      "step": 895
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3144020736217499,
      "learning_rate": 0.0006654901590725577,
      "loss": 2.462,
      "step": 896
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3034922182559967,
      "learning_rate": 0.0006647888117065507,
      "loss": 2.3567,
      "step": 897
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3094896376132965,
      "learning_rate": 0.0006640871004109086,
      "loss": 2.4119,
      "step": 898
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.2995099127292633,
      "learning_rate": 0.000663385026735334,
      "loss": 2.4437,
      "step": 899
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3217426836490631,
      "learning_rate": 0.0006626825922303285,
      "loss": 2.4282,
      "step": 900
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.32670843601226807,
      "learning_rate": 0.0006619797984471915,
      "loss": 2.3024,
      "step": 901
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3343380391597748,
      "learning_rate": 0.0006612766469380158,
      "loss": 2.3918,
      "step": 902
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3268018662929535,
      "learning_rate": 0.0006605731392556833,
      "loss": 2.3906,
      "step": 903
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3148239254951477,
      "learning_rate": 0.0006598692769538637,
      "loss": 2.4497,
      "step": 904
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.31122663617134094,
      "learning_rate": 0.0006591650615870091,
      "loss": 2.3482,
      "step": 905
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.31533190608024597,
      "learning_rate": 0.0006584604947103514,
      "loss": 2.295,
      "step": 906
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.28828954696655273,
      "learning_rate": 0.0006577555778798993,
      "loss": 2.3912,
      "step": 907
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.31450262665748596,
      "learning_rate": 0.0006570503126524336,
      "loss": 2.3851,
      "step": 908
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2936091721057892,
      "learning_rate": 0.0006563447005855054,
      "loss": 2.3953,
      "step": 909
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3030509352684021,
      "learning_rate": 0.000655638743237431,
      "loss": 2.4061,
      "step": 910
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.32506605982780457,
      "learning_rate": 0.0006549324421672894,
      "loss": 2.4283,
      "step": 911
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3360969126224518,
      "learning_rate": 0.0006542257989349194,
      "loss": 2.2549,
      "step": 912
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.31918394565582275,
      "learning_rate": 0.0006535188151009142,
      "loss": 2.2187,
      "step": 913
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3269701600074768,
      "learning_rate": 0.0006528114922266204,
      "loss": 2.3891,
      "step": 914
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.30365926027297974,
      "learning_rate": 0.0006521038318741327,
      "loss": 2.4356,
      "step": 915
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.30986249446868896,
      "learning_rate": 0.0006513958356062912,
      "loss": 2.401,
      "step": 916
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3011469542980194,
      "learning_rate": 0.0006506875049866781,
      "loss": 2.4734,
      "step": 917
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.29565685987472534,
      "learning_rate": 0.0006499788415796137,
      "loss": 2.5412,
      "step": 918
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2885633111000061,
      "learning_rate": 0.0006492698469501532,
      "loss": 2.3397,
      "step": 919
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.29744571447372437,
      "learning_rate": 0.0006485605226640837,
      "loss": 2.4436,
      "step": 920
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.29583558440208435,
      "learning_rate": 0.00064785087028792,
      "loss": 2.4449,
      "step": 921
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2923334836959839,
      "learning_rate": 0.0006471408913889019,
      "loss": 2.3371,
      "step": 922
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.287663072347641,
      "learning_rate": 0.0006464305875349892,
      "loss": 2.4026,
      "step": 923
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2928681969642639,
      "learning_rate": 0.000645719960294861,
      "loss": 2.3949,
      "step": 924
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2832850217819214,
      "learning_rate": 0.0006450090112379092,
      "loss": 2.302,
      "step": 925
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3099982440471649,
      "learning_rate": 0.0006442977419342371,
      "loss": 2.4761,
      "step": 926
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3094823658466339,
      "learning_rate": 0.000643586153954655,
      "loss": 2.3866,
      "step": 927
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2930482029914856,
      "learning_rate": 0.0006428742488706772,
      "loss": 2.3602,
      "step": 928
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.32160472869873047,
      "learning_rate": 0.0006421620282545182,
      "loss": 2.3782,
      "step": 929
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3018878698348999,
      "learning_rate": 0.0006414494936790892,
      "loss": 2.3259,
      "step": 930
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2942325472831726,
      "learning_rate": 0.0006407366467179951,
      "loss": 2.2231,
      "step": 931
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3061409294605255,
      "learning_rate": 0.0006400234889455301,
      "loss": 2.3928,
      "step": 932
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.31030675768852234,
      "learning_rate": 0.0006393100219366755,
      "loss": 2.378,
      "step": 933
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.32317283749580383,
      "learning_rate": 0.0006385962472670953,
      "loss": 2.3409,
      "step": 934
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3387592136859894,
      "learning_rate": 0.0006378821665131328,
      "loss": 2.5154,
      "step": 935
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.31696566939353943,
      "learning_rate": 0.0006371677812518072,
      "loss": 2.423,
      "step": 936
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.30330392718315125,
      "learning_rate": 0.0006364530930608107,
      "loss": 2.3365,
      "step": 937
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.29808762669563293,
      "learning_rate": 0.0006357381035185038,
      "loss": 2.2449,
      "step": 938
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.31194189190864563,
      "learning_rate": 0.0006350228142039131,
      "loss": 2.3091,
      "step": 939
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.30580538511276245,
      "learning_rate": 0.000634307226696727,
      "loss": 2.4408,
      "step": 940
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.296090304851532,
      "learning_rate": 0.0006335913425772926,
      "loss": 2.3478,
      "step": 941
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2944009602069855,
      "learning_rate": 0.0006328751634266117,
      "loss": 2.4011,
      "step": 942
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.29752057790756226,
      "learning_rate": 0.0006321586908263382,
      "loss": 2.3208,
      "step": 943
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3037665784358978,
      "learning_rate": 0.0006314419263587732,
      "loss": 2.448,
      "step": 944
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3061296343803406,
      "learning_rate": 0.0006307248716068637,
      "loss": 2.3482,
      "step": 945
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.29839545488357544,
      "learning_rate": 0.0006300075281541964,
      "loss": 2.3299,
      "step": 946
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3166232109069824,
      "learning_rate": 0.0006292898975849966,
      "loss": 2.3348,
      "step": 947
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.30298155546188354,
      "learning_rate": 0.000628571981484123,
      "loss": 2.2931,
      "step": 948
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3170483410358429,
      "learning_rate": 0.0006278537814370654,
      "loss": 2.3583,
      "step": 949
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.32711687684059143,
      "learning_rate": 0.0006271352990299406,
      "loss": 2.438,
      "step": 950
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.31328675150871277,
      "learning_rate": 0.0006264165358494885,
      "loss": 2.3887,
      "step": 951
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3098021149635315,
      "learning_rate": 0.0006256974934830694,
      "loss": 2.4755,
      "step": 952
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.29550814628601074,
      "learning_rate": 0.0006249781735186606,
      "loss": 2.3933,
      "step": 953
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.31168827414512634,
      "learning_rate": 0.0006242585775448518,
      "loss": 2.366,
      "step": 954
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.30768534541130066,
      "learning_rate": 0.0006235387071508427,
      "loss": 2.2886,
      "step": 955
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3338025212287903,
      "learning_rate": 0.0006228185639264384,
      "loss": 2.443,
      "step": 956
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3087279498577118,
      "learning_rate": 0.0006220981494620475,
      "loss": 2.388,
      "step": 957
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.321492075920105,
      "learning_rate": 0.000621377465348677,
      "loss": 2.4613,
      "step": 958
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.31625381112098694,
      "learning_rate": 0.0006206565131779293,
      "loss": 2.3744,
      "step": 959
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.29133936762809753,
      "learning_rate": 0.0006199352945419994,
      "loss": 2.4568,
      "step": 960
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.32993292808532715,
      "learning_rate": 0.00061921381103367,
      "loss": 2.3629,
      "step": 961
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.31235793232917786,
      "learning_rate": 0.0006184920642463094,
      "loss": 2.4127,
      "step": 962
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.32164865732192993,
      "learning_rate": 0.0006177700557738672,
      "loss": 2.4418,
      "step": 963
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.30301621556282043,
      "learning_rate": 0.0006170477872108706,
      "loss": 2.2264,
      "step": 964
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.319734662771225,
      "learning_rate": 0.0006163252601524216,
      "loss": 2.4682,
      "step": 965
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3308180570602417,
      "learning_rate": 0.0006156024761941925,
      "loss": 2.3563,
      "step": 966
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3379901647567749,
      "learning_rate": 0.000614879436932424,
      "loss": 2.4432,
      "step": 967
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.30172112584114075,
      "learning_rate": 0.0006141561439639196,
      "loss": 2.4625,
      "step": 968
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.30462685227394104,
      "learning_rate": 0.0006134325988860433,
      "loss": 2.382,
      "step": 969
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3028826117515564,
      "learning_rate": 0.0006127088032967165,
      "loss": 2.4978,
      "step": 970
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.30195313692092896,
      "learning_rate": 0.0006119847587944131,
      "loss": 2.3465,
      "step": 971
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2951676547527313,
      "learning_rate": 0.0006112604669781572,
      "loss": 2.3348,
      "step": 972
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.29948189854621887,
      "learning_rate": 0.0006105359294475188,
      "loss": 2.4519,
      "step": 973
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.29389581084251404,
      "learning_rate": 0.0006098111478026107,
      "loss": 2.4557,
      "step": 974
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3002973198890686,
      "learning_rate": 0.0006090861236440848,
      "loss": 2.3619,
      "step": 975
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.28676360845565796,
      "learning_rate": 0.0006083608585731282,
      "loss": 2.3322,
      "step": 976
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3026236295700073,
      "learning_rate": 0.0006076353541914609,
      "loss": 2.3527,
      "step": 977
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.29628345370292664,
      "learning_rate": 0.0006069096121013307,
      "loss": 2.3516,
      "step": 978
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3044242262840271,
      "learning_rate": 0.0006061836339055105,
      "loss": 2.3244,
      "step": 979
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3011431396007538,
      "learning_rate": 0.0006054574212072948,
      "loss": 2.2927,
      "step": 980
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3243943750858307,
      "learning_rate": 0.0006047309756104958,
      "loss": 2.4018,
      "step": 981
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3188526928424835,
      "learning_rate": 0.00060400429871944,
      "loss": 2.3906,
      "step": 982
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.318202406167984,
      "learning_rate": 0.0006032773921389654,
      "loss": 2.4639,
      "step": 983
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3064056932926178,
      "learning_rate": 0.0006025502574744162,
      "loss": 2.413,
      "step": 984
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.292248398065567,
      "learning_rate": 0.000601822896331641,
      "loss": 2.2977,
      "step": 985
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.27767884731292725,
      "learning_rate": 0.0006010953103169883,
      "loss": 2.3019,
      "step": 986
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2944405972957611,
      "learning_rate": 0.0006003675010373034,
      "loss": 2.4127,
      "step": 987
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2957027554512024,
      "learning_rate": 0.0005996394700999246,
      "loss": 2.3319,
      "step": 988
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3164874315261841,
      "learning_rate": 0.0005989112191126794,
      "loss": 2.4625,
      "step": 989
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3009915053844452,
      "learning_rate": 0.0005981827496838822,
      "loss": 2.4432,
      "step": 990
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2985946238040924,
      "learning_rate": 0.0005974540634223286,
      "loss": 2.3582,
      "step": 991
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3122430741786957,
      "learning_rate": 0.0005967251619372939,
      "loss": 2.3846,
      "step": 992
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.3289636969566345,
      "learning_rate": 0.0005959960468385284,
      "loss": 2.3728,
      "step": 993
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.30823132395744324,
      "learning_rate": 0.0005952667197362542,
      "loss": 2.3509,
      "step": 994
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.30790138244628906,
      "learning_rate": 0.0005945371822411621,
      "loss": 2.3739,
      "step": 995
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.31040656566619873,
      "learning_rate": 0.0005938074359644063,
      "loss": 2.2972,
      "step": 996
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.31253504753112793,
      "learning_rate": 0.0005930774825176034,
      "loss": 2.3479,
      "step": 997
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.29523923993110657,
      "learning_rate": 0.0005923473235128268,
      "loss": 2.2666,
      "step": 998
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.30329111218452454,
      "learning_rate": 0.0005916169605626042,
      "loss": 2.3963,
      "step": 999
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.300688236951828,
      "learning_rate": 0.0005908863952799134,
      "loss": 2.3005,
      "step": 1000
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.3208443820476532,
      "learning_rate": 0.0005901556292781793,
      "loss": 2.3411,
      "step": 1001
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.298960417509079,
      "learning_rate": 0.0005894246641712698,
      "loss": 2.3577,
      "step": 1002
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.31206604838371277,
      "learning_rate": 0.0005886935015734931,
      "loss": 2.4122,
      "step": 1003
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.3135763108730316,
      "learning_rate": 0.0005879621430995928,
      "loss": 2.3315,
      "step": 1004
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.31600093841552734,
      "learning_rate": 0.0005872305903647455,
      "loss": 2.3668,
      "step": 1005
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.30196017026901245,
      "learning_rate": 0.0005864988449845569,
      "loss": 2.4128,
      "step": 1006
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.306643009185791,
      "learning_rate": 0.0005857669085750578,
      "loss": 2.3594,
      "step": 1007
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.30554085969924927,
      "learning_rate": 0.0005850347827527013,
      "loss": 2.3017,
      "step": 1008
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.29915153980255127,
      "learning_rate": 0.0005843024691343584,
      "loss": 2.3422,
      "step": 1009
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.30357787013053894,
      "learning_rate": 0.000583569969337315,
      "loss": 2.4172,
      "step": 1010
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.32609450817108154,
      "learning_rate": 0.0005828372849792686,
      "loss": 2.4061,
      "step": 1011
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.32521572709083557,
      "learning_rate": 0.0005821044176783234,
      "loss": 2.4376,
      "step": 1012
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2943490147590637,
      "learning_rate": 0.0005813713690529886,
      "loss": 2.4286,
      "step": 1013
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3063652515411377,
      "learning_rate": 0.0005806381407221729,
      "loss": 2.3001,
      "step": 1014
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3061884343624115,
      "learning_rate": 0.0005799047343051826,
      "loss": 2.3186,
      "step": 1015
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.29832541942596436,
      "learning_rate": 0.0005791711514217171,
      "loss": 2.3981,
      "step": 1016
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3024725914001465,
      "learning_rate": 0.0005784373936918654,
      "loss": 2.3661,
      "step": 1017
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2888949513435364,
      "learning_rate": 0.0005777034627361025,
      "loss": 2.3488,
      "step": 1018
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3165346086025238,
      "learning_rate": 0.0005769693601752864,
      "loss": 2.4181,
      "step": 1019
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.30915242433547974,
      "learning_rate": 0.0005762350876306537,
      "loss": 2.3472,
      "step": 1020
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3151189088821411,
      "learning_rate": 0.0005755006467238168,
      "loss": 2.4454,
      "step": 1021
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.30571332573890686,
      "learning_rate": 0.0005747660390767593,
      "loss": 2.4168,
      "step": 1022
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.31015849113464355,
      "learning_rate": 0.0005740312663118338,
      "loss": 2.3569,
      "step": 1023
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.29920196533203125,
      "learning_rate": 0.0005732963300517568,
      "loss": 2.3311,
      "step": 1024
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.29385101795196533,
      "learning_rate": 0.0005725612319196064,
      "loss": 2.4137,
      "step": 1025
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.31059563159942627,
      "learning_rate": 0.000571825973538818,
      "loss": 2.4112,
      "step": 1026
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2965165674686432,
      "learning_rate": 0.0005710905565331811,
      "loss": 2.3716,
      "step": 1027
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3007526993751526,
      "learning_rate": 0.0005703549825268353,
      "loss": 2.4326,
      "step": 1028
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.33964696526527405,
      "learning_rate": 0.0005696192531442667,
      "loss": 2.2581,
      "step": 1029
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.30018141865730286,
      "learning_rate": 0.000568883370010305,
      "loss": 2.425,
      "step": 1030
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.30151447653770447,
      "learning_rate": 0.0005681473347501192,
      "loss": 2.2983,
      "step": 1031
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.309883713722229,
      "learning_rate": 0.0005674111489892144,
      "loss": 2.3406,
      "step": 1032
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3318580985069275,
      "learning_rate": 0.0005666748143534282,
      "loss": 2.3004,
      "step": 1033
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.33882084488868713,
      "learning_rate": 0.0005659383324689266,
      "loss": 2.3626,
      "step": 1034
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.29993513226509094,
      "learning_rate": 0.0005652017049622007,
      "loss": 2.3967,
      "step": 1035
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.310144305229187,
      "learning_rate": 0.0005644649334600641,
      "loss": 2.3397,
      "step": 1036
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.30329033732414246,
      "learning_rate": 0.0005637280195896474,
      "loss": 2.3289,
      "step": 1037
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3277477025985718,
      "learning_rate": 0.0005629909649783961,
      "loss": 2.286,
      "step": 1038
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.31896236538887024,
      "learning_rate": 0.0005622537712540664,
      "loss": 2.3851,
      "step": 1039
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2991380989551544,
      "learning_rate": 0.0005615164400447218,
      "loss": 2.3184,
      "step": 1040
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2978149652481079,
      "learning_rate": 0.0005607789729787294,
      "loss": 2.4227,
      "step": 1041
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.32530879974365234,
      "learning_rate": 0.0005600413716847564,
      "loss": 2.3837,
      "step": 1042
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3316652476787567,
      "learning_rate": 0.000559303637791766,
      "loss": 2.4506,
      "step": 1043
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.32308170199394226,
      "learning_rate": 0.0005585657729290151,
      "loss": 2.3254,
      "step": 1044
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.30372336506843567,
      "learning_rate": 0.000557827778726049,
      "loss": 2.3908,
      "step": 1045
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2876874506473541,
      "learning_rate": 0.0005570896568126993,
      "loss": 2.3302,
      "step": 1046
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3253418207168579,
      "learning_rate": 0.0005563514088190788,
      "loss": 2.3099,
      "step": 1047
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3016171157360077,
      "learning_rate": 0.0005556130363755798,
      "loss": 2.4142,
      "step": 1048
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2951292395591736,
      "learning_rate": 0.0005548745411128688,
      "loss": 2.3406,
      "step": 1049
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.30991533398628235,
      "learning_rate": 0.0005541359246618835,
      "loss": 2.4021,
      "step": 1050
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.29751136898994446,
      "learning_rate": 0.0005533971886538293,
      "loss": 2.3352,
      "step": 1051
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3051947355270386,
      "learning_rate": 0.000552658334720176,
      "loss": 2.3637,
      "step": 1052
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2904837429523468,
      "learning_rate": 0.0005519193644926535,
      "loss": 2.2998,
      "step": 1053
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.32340672612190247,
      "learning_rate": 0.0005511802796032485,
      "loss": 2.4111,
      "step": 1054
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3361344337463379,
      "learning_rate": 0.0005504410816842009,
      "loss": 2.4288,
      "step": 1055
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.307147353887558,
      "learning_rate": 0.0005497017723680009,
      "loss": 2.3666,
      "step": 1056
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2846252918243408,
      "learning_rate": 0.0005489623532873836,
      "loss": 2.3468,
      "step": 1057
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3021845519542694,
      "learning_rate": 0.0005482228260753273,
      "loss": 2.3812,
      "step": 1058
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.28622153401374817,
      "learning_rate": 0.0005474831923650488,
      "loss": 2.4421,
      "step": 1059
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3023550808429718,
      "learning_rate": 0.00054674345379,
      "loss": 2.3262,
      "step": 1060
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2916857600212097,
      "learning_rate": 0.000546003611983865,
      "loss": 2.3693,
      "step": 1061
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.30066028237342834,
      "learning_rate": 0.0005452636685805552,
      "loss": 2.4003,
      "step": 1062
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3068836033344269,
      "learning_rate": 0.0005445236252142066,
      "loss": 2.2854,
      "step": 1063
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3259165287017822,
      "learning_rate": 0.000543783483519176,
      "loss": 2.3529,
      "step": 1064
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2854894995689392,
      "learning_rate": 0.0005430432451300374,
      "loss": 2.4138,
      "step": 1065
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.29939308762550354,
      "learning_rate": 0.0005423029116815781,
      "loss": 2.358,
      "step": 1066
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3152993321418762,
      "learning_rate": 0.0005415624848087959,
      "loss": 2.3082,
      "step": 1067
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2994430363178253,
      "learning_rate": 0.000540821966146894,
      "loss": 2.3926,
      "step": 1068
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.30943217873573303,
      "learning_rate": 0.0005400813573312793,
      "loss": 2.3989,
      "step": 1069
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.30771803855895996,
      "learning_rate": 0.0005393406599975572,
      "loss": 2.3013,
      "step": 1070
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2922094762325287,
      "learning_rate": 0.0005385998757815287,
      "loss": 2.4235,
      "step": 1071
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2924877405166626,
      "learning_rate": 0.0005378590063191867,
      "loss": 2.3746,
      "step": 1072
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.27533644437789917,
      "learning_rate": 0.0005371180532467124,
      "loss": 2.3071,
      "step": 1073
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.30002379417419434,
      "learning_rate": 0.000536377018200472,
      "loss": 2.3481,
      "step": 1074
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2966727614402771,
      "learning_rate": 0.0005356359028170118,
      "loss": 2.3734,
      "step": 1075
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.31297966837882996,
      "learning_rate": 0.0005348947087330564,
      "loss": 2.4367,
      "step": 1076
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3101723790168762,
      "learning_rate": 0.0005341534375855037,
      "loss": 2.3558,
      "step": 1077
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.29059550166130066,
      "learning_rate": 0.0005334120910114222,
      "loss": 2.3193,
      "step": 1078
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3061675429344177,
      "learning_rate": 0.0005326706706480467,
      "loss": 2.4138,
      "step": 1079
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.30652984976768494,
      "learning_rate": 0.0005319291781327749,
      "loss": 2.3405,
      "step": 1080
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3158240020275116,
      "learning_rate": 0.0005311876151031642,
      "loss": 2.3113,
      "step": 1081
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3115432560443878,
      "learning_rate": 0.0005304459831969274,
      "loss": 2.2932,
      "step": 1082
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.30872443318367004,
      "learning_rate": 0.0005297042840519294,
      "loss": 2.3517,
      "step": 1083
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.32354041934013367,
      "learning_rate": 0.0005289625193061838,
      "loss": 2.3927,
      "step": 1084
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.30163508653640747,
      "learning_rate": 0.0005282206905978489,
      "loss": 2.3615,
      "step": 1085
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.31691834330558777,
      "learning_rate": 0.0005274787995652246,
      "loss": 2.3552,
      "step": 1086
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.30992361903190613,
      "learning_rate": 0.000526736847846748,
      "loss": 2.3653,
      "step": 1087
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3007590174674988,
      "learning_rate": 0.0005259948370809901,
      "loss": 2.3045,
      "step": 1088
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.31005585193634033,
      "learning_rate": 0.0005252527689066533,
      "loss": 2.3382,
      "step": 1089
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3091129958629608,
      "learning_rate": 0.0005245106449625654,
      "loss": 2.4162,
      "step": 1090
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.29265978932380676,
      "learning_rate": 0.0005237684668876785,
      "loss": 2.3964,
      "step": 1091
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.29794469475746155,
      "learning_rate": 0.0005230262363210637,
      "loss": 2.3023,
      "step": 1092
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.29068028926849365,
      "learning_rate": 0.0005222839549019079,
      "loss": 2.3341,
      "step": 1093
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2966498136520386,
      "learning_rate": 0.0005215416242695108,
      "loss": 2.3291,
      "step": 1094
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3571353554725647,
      "learning_rate": 0.0005207992460632804,
      "loss": 2.3052,
      "step": 1095
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.31969648599624634,
      "learning_rate": 0.0005200568219227299,
      "loss": 2.3789,
      "step": 1096
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3139449656009674,
      "learning_rate": 0.000519314353487474,
      "loss": 2.4639,
      "step": 1097
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.30797702074050903,
      "learning_rate": 0.0005185718423972251,
      "loss": 2.3164,
      "step": 1098
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.32974007725715637,
      "learning_rate": 0.0005178292902917898,
      "loss": 2.3896,
      "step": 1099
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.292894572019577,
      "learning_rate": 0.0005170866988110656,
      "loss": 2.232,
      "step": 1100
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.31161773204803467,
      "learning_rate": 0.0005163440695950362,
      "loss": 2.4,
      "step": 1101
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.3129146099090576,
      "learning_rate": 0.0005156014042837695,
      "loss": 2.3265,
      "step": 1102
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.28752321004867554,
      "learning_rate": 0.0005148587045174128,
      "loss": 2.2989,
      "step": 1103
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2952859699726105,
      "learning_rate": 0.0005141159719361891,
      "loss": 2.3742,
      "step": 1104
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.3028091490268707,
      "learning_rate": 0.0005133732081803945,
      "loss": 2.2794,
      "step": 1105
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.29412439465522766,
      "learning_rate": 0.0005126304148903936,
      "loss": 2.324,
      "step": 1106
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.30290886759757996,
      "learning_rate": 0.0005118875937066161,
      "loss": 2.3274,
      "step": 1107
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2967037260532379,
      "learning_rate": 0.0005111447462695537,
      "loss": 2.3576,
      "step": 1108
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.29620665311813354,
      "learning_rate": 0.0005104018742197557,
      "loss": 2.2641,
      "step": 1109
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2882354259490967,
      "learning_rate": 0.0005096589791978261,
      "loss": 2.2816,
      "step": 1110
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2922405004501343,
      "learning_rate": 0.0005089160628444192,
      "loss": 2.3211,
      "step": 1111
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.3217419683933258,
      "learning_rate": 0.0005081731268002371,
      "loss": 2.3231,
      "step": 1112
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.3034839928150177,
      "learning_rate": 0.0005074301727060243,
      "loss": 2.4277,
      "step": 1113
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.30176806449890137,
      "learning_rate": 0.0005066872022025663,
      "loss": 2.3583,
      "step": 1114
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.289810448884964,
      "learning_rate": 0.0005059442169306844,
      "loss": 2.3358,
      "step": 1115
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.287119060754776,
      "learning_rate": 0.0005052012185312321,
      "loss": 2.261,
      "step": 1116
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.305807501077652,
      "learning_rate": 0.0005044582086450925,
      "loss": 2.3205,
      "step": 1117
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.31006693840026855,
      "learning_rate": 0.0005037151889131737,
      "loss": 2.3058,
      "step": 1118
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.31865015625953674,
      "learning_rate": 0.0005029721609764059,
      "loss": 2.3497,
      "step": 1119
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.30271947383880615,
      "learning_rate": 0.000502229126475737,
      "loss": 2.3472,
      "step": 1120
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.3244991600513458,
      "learning_rate": 0.0005014860870521293,
      "loss": 2.3157,
      "step": 1121
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2881084978580475,
      "learning_rate": 0.0005007430443465569,
      "loss": 2.27,
      "step": 1122
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2968061566352844,
      "learning_rate": 0.0005,
      "loss": 2.3123,
      "step": 1123
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.31666702032089233,
      "learning_rate": 0.0004992569556534432,
      "loss": 2.4076,
      "step": 1124
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3246583342552185,
      "learning_rate": 0.0004985139129478707,
      "loss": 2.3002,
      "step": 1125
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2949075698852539,
      "learning_rate": 0.0004977708735242633,
      "loss": 2.3889,
      "step": 1126
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3042008876800537,
      "learning_rate": 0.0004970278390235942,
      "loss": 2.3184,
      "step": 1127
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2993326485157013,
      "learning_rate": 0.0004962848110868262,
      "loss": 2.2718,
      "step": 1128
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3096159100532532,
      "learning_rate": 0.0004955417913549074,
      "loss": 2.3593,
      "step": 1129
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.29892459511756897,
      "learning_rate": 0.0004947987814687679,
      "loss": 2.314,
      "step": 1130
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3247067630290985,
      "learning_rate": 0.0004940557830693157,
      "loss": 2.4382,
      "step": 1131
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3024771511554718,
      "learning_rate": 0.0004933127977974338,
      "loss": 2.4153,
      "step": 1132
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.28963401913642883,
      "learning_rate": 0.0004925698272939757,
      "loss": 2.337,
      "step": 1133
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.29711440205574036,
      "learning_rate": 0.0004918268731997632,
      "loss": 2.2698,
      "step": 1134
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3108326196670532,
      "learning_rate": 0.0004910839371555809,
      "loss": 2.3548,
      "step": 1135
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3086743652820587,
      "learning_rate": 0.0004903410208021739,
      "loss": 2.3817,
      "step": 1136
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3094700872898102,
      "learning_rate": 0.0004895981257802443,
      "loss": 2.3877,
      "step": 1137
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.30902403593063354,
      "learning_rate": 0.0004888552537304463,
      "loss": 2.4169,
      "step": 1138
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3139355480670929,
      "learning_rate": 0.00048811240629338394,
      "loss": 2.339,
      "step": 1139
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.30372169613838196,
      "learning_rate": 0.00048736958510960663,
      "loss": 2.4416,
      "step": 1140
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.30429643392562866,
      "learning_rate": 0.00048662679181960564,
      "loss": 2.2063,
      "step": 1141
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.29687902331352234,
      "learning_rate": 0.00048588402806381094,
      "loss": 2.3516,
      "step": 1142
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2939431965351105,
      "learning_rate": 0.0004851412954825874,
      "loss": 2.3974,
      "step": 1143
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2931225299835205,
      "learning_rate": 0.00048439859571623034,
      "loss": 2.3577,
      "step": 1144
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.29703518748283386,
      "learning_rate": 0.00048365593040496373,
      "loss": 2.3283,
      "step": 1145
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.28708407282829285,
      "learning_rate": 0.00048291330118893443,
      "loss": 2.3838,
      "step": 1146
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2889479994773865,
      "learning_rate": 0.0004821707097082102,
      "loss": 2.3371,
      "step": 1147
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.30706197023391724,
      "learning_rate": 0.0004814281576027749,
      "loss": 2.3521,
      "step": 1148
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.30677467584609985,
      "learning_rate": 0.000480685646512526,
      "loss": 2.3098,
      "step": 1149
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.3017318844795227,
      "learning_rate": 0.00047994317807727025,
      "loss": 2.3743,
      "step": 1150
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2797110974788666,
      "learning_rate": 0.00047920075393671974,
      "loss": 2.3066,
      "step": 1151
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.32897502183914185,
      "learning_rate": 0.0004784583757304893,
      "loss": 2.4055,
      "step": 1152
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.30617067217826843,
      "learning_rate": 0.00047771604509809214,
      "loss": 2.4238,
      "step": 1153
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.30441275238990784,
      "learning_rate": 0.0004769737636789364,
      "loss": 2.4218,
      "step": 1154
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.3035639524459839,
      "learning_rate": 0.00047623153311232157,
      "loss": 2.405,
      "step": 1155
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.3006420433521271,
      "learning_rate": 0.0004754893550374346,
      "loss": 2.3763,
      "step": 1156
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.28726086020469666,
      "learning_rate": 0.00047474723109334685,
      "loss": 2.251,
      "step": 1157
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.30908524990081787,
      "learning_rate": 0.00047400516291900993,
      "loss": 2.2817,
      "step": 1158
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.3173764646053314,
      "learning_rate": 0.0004732631521532522,
      "loss": 2.3145,
      "step": 1159
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.3071606457233429,
      "learning_rate": 0.0004725212004347755,
      "loss": 2.3323,
      "step": 1160
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.30247968435287476,
      "learning_rate": 0.00047177930940215095,
      "loss": 2.2113,
      "step": 1161
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.31214964389801025,
      "learning_rate": 0.00047103748069381624,
      "loss": 2.2446,
      "step": 1162
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.29122015833854675,
      "learning_rate": 0.0004702957159480707,
      "loss": 2.2731,
      "step": 1163
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2985371947288513,
      "learning_rate": 0.00046955401680307267,
      "loss": 2.3177,
      "step": 1164
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.3124462962150574,
      "learning_rate": 0.0004688123848968359,
      "loss": 2.3448,
      "step": 1165
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.31831255555152893,
      "learning_rate": 0.00046807082186722516,
      "loss": 2.3161,
      "step": 1166
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3008701503276825,
      "learning_rate": 0.0004673293293519535,
      "loss": 2.2989,
      "step": 1167
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3007342219352722,
      "learning_rate": 0.00046658790898857806,
      "loss": 2.3616,
      "step": 1168
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3050384819507599,
      "learning_rate": 0.0004658465624144963,
      "loss": 2.2816,
      "step": 1169
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.310861736536026,
      "learning_rate": 0.0004651052912669438,
      "loss": 2.3776,
      "step": 1170
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2897826135158539,
      "learning_rate": 0.0004643640971829883,
      "loss": 2.444,
      "step": 1171
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.30053508281707764,
      "learning_rate": 0.0004636229817995281,
      "loss": 2.2317,
      "step": 1172
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.29412463307380676,
      "learning_rate": 0.0004628819467532876,
      "loss": 2.2298,
      "step": 1173
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.30548563599586487,
      "learning_rate": 0.00046214099368081335,
      "loss": 2.3898,
      "step": 1174
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2910051643848419,
      "learning_rate": 0.0004614001242184714,
      "loss": 2.2721,
      "step": 1175
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3035147786140442,
      "learning_rate": 0.000460659340002443,
      "loss": 2.302,
      "step": 1176
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.31011316180229187,
      "learning_rate": 0.00045991864266872073,
      "loss": 2.37,
      "step": 1177
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.30583709478378296,
      "learning_rate": 0.00045917803385310595,
      "loss": 2.3636,
      "step": 1178
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.5577784776687622,
      "learning_rate": 0.00045843751519120417,
      "loss": 2.3413,
      "step": 1179
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.30908486247062683,
      "learning_rate": 0.00045769708831842193,
      "loss": 2.369,
      "step": 1180
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2877659499645233,
      "learning_rate": 0.00045695675486996266,
      "loss": 2.2912,
      "step": 1181
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.31634849309921265,
      "learning_rate": 0.00045621651648082405,
      "loss": 2.3209,
      "step": 1182
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3229602575302124,
      "learning_rate": 0.00045547637478579356,
      "loss": 2.2827,
      "step": 1183
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.30948737263679504,
      "learning_rate": 0.0004547363314194449,
      "loss": 2.2872,
      "step": 1184
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3113803565502167,
      "learning_rate": 0.000453996388016135,
      "loss": 2.3875,
      "step": 1185
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.324492484331131,
      "learning_rate": 0.0004532565462099999,
      "loss": 2.3214,
      "step": 1186
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.30291733145713806,
      "learning_rate": 0.0004525168076349513,
      "loss": 2.3474,
      "step": 1187
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3210369050502777,
      "learning_rate": 0.0004517771739246729,
      "loss": 2.4375,
      "step": 1188
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.29424765706062317,
      "learning_rate": 0.0004510376467126165,
      "loss": 2.3305,
      "step": 1189
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2933778762817383,
      "learning_rate": 0.0004502982276319992,
      "loss": 2.334,
      "step": 1190
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2952924072742462,
      "learning_rate": 0.0004495589183157991,
      "loss": 2.3423,
      "step": 1191
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3097077012062073,
      "learning_rate": 0.0004488197203967517,
      "loss": 2.3092,
      "step": 1192
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2843981683254242,
      "learning_rate": 0.0004480806355073467,
      "loss": 2.3745,
      "step": 1193
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3039379417896271,
      "learning_rate": 0.000447341665279824,
      "loss": 2.3195,
      "step": 1194
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.31840455532073975,
      "learning_rate": 0.0004466028113461708,
      "loss": 2.3654,
      "step": 1195
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3066041171550751,
      "learning_rate": 0.0004458640753381167,
      "loss": 2.4601,
      "step": 1196
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2985830008983612,
      "learning_rate": 0.0004451254588871313,
      "loss": 2.2966,
      "step": 1197
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3200114071369171,
      "learning_rate": 0.0004443869636244203,
      "loss": 2.3506,
      "step": 1198
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.31671223044395447,
      "learning_rate": 0.0004436485911809212,
      "loss": 2.3213,
      "step": 1199
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3175174295902252,
      "learning_rate": 0.00044291034318730087,
      "loss": 2.3538,
      "step": 1200
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3200148940086365,
      "learning_rate": 0.0004421722212739511,
      "loss": 2.3547,
      "step": 1201
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.29213330149650574,
      "learning_rate": 0.0004414342270709848,
      "loss": 2.3247,
      "step": 1202
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3207609951496124,
      "learning_rate": 0.00044069636220823397,
      "loss": 2.3523,
      "step": 1203
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.28984445333480835,
      "learning_rate": 0.0004399586283152437,
      "loss": 2.3353,
      "step": 1204
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2810138463973999,
      "learning_rate": 0.0004392210270212706,
      "loss": 2.3424,
      "step": 1205
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.29400312900543213,
      "learning_rate": 0.00043848355995527825,
      "loss": 2.2652,
      "step": 1206
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2941107451915741,
      "learning_rate": 0.00043774622874593374,
      "loss": 2.2771,
      "step": 1207
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2921290099620819,
      "learning_rate": 0.000437009035021604,
      "loss": 2.4171,
      "step": 1208
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2998519241809845,
      "learning_rate": 0.00043627198041035274,
      "loss": 2.3385,
      "step": 1209
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.29157644510269165,
      "learning_rate": 0.00043553506653993597,
      "loss": 2.278,
      "step": 1210
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3100161552429199,
      "learning_rate": 0.0004347982950377992,
      "loss": 2.3367,
      "step": 1211
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2907159924507141,
      "learning_rate": 0.0004340616675310735,
      "loss": 2.3676,
      "step": 1212
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3012669086456299,
      "learning_rate": 0.00043332518564657193,
      "loss": 2.2593,
      "step": 1213
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2989492118358612,
      "learning_rate": 0.0004325888510107856,
      "loss": 2.2311,
      "step": 1214
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.30211102962493896,
      "learning_rate": 0.0004318526652498809,
      "loss": 2.3854,
      "step": 1215
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2904473543167114,
      "learning_rate": 0.00043111662998969523,
      "loss": 2.3328,
      "step": 1216
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.30228090286254883,
      "learning_rate": 0.0004303807468557335,
      "loss": 2.4172,
      "step": 1217
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2916053831577301,
      "learning_rate": 0.0004296450174731648,
      "loss": 2.3308,
      "step": 1218
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2973944842815399,
      "learning_rate": 0.0004289094434668188,
      "loss": 2.3179,
      "step": 1219
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2977837324142456,
      "learning_rate": 0.00042817402646118185,
      "loss": 2.3262,
      "step": 1220
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2889993190765381,
      "learning_rate": 0.0004274387680803936,
      "loss": 2.2938,
      "step": 1221
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.30136722326278687,
      "learning_rate": 0.00042670366994824327,
      "loss": 2.2388,
      "step": 1222
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.31258583068847656,
      "learning_rate": 0.0004259687336881663,
      "loss": 2.4132,
      "step": 1223
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3046659827232361,
      "learning_rate": 0.0004252339609232408,
      "loss": 2.2966,
      "step": 1224
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3157033324241638,
      "learning_rate": 0.0004244993532761834,
      "loss": 2.2675,
      "step": 1225
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3054787516593933,
      "learning_rate": 0.00042376491236934634,
      "loss": 2.2896,
      "step": 1226
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.29731202125549316,
      "learning_rate": 0.0004230306398247136,
      "loss": 2.4027,
      "step": 1227
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3012170195579529,
      "learning_rate": 0.0004222965372638976,
      "loss": 2.3227,
      "step": 1228
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.311450719833374,
      "learning_rate": 0.0004215626063081348,
      "loss": 2.4008,
      "step": 1229
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.30786025524139404,
      "learning_rate": 0.000420828848578283,
      "loss": 2.3319,
      "step": 1230
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.29442715644836426,
      "learning_rate": 0.0004200952656948175,
      "loss": 2.2715,
      "step": 1231
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2859747111797333,
      "learning_rate": 0.0004193618592778272,
      "loss": 2.4024,
      "step": 1232
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.28379589319229126,
      "learning_rate": 0.0004186286309470116,
      "loss": 2.3972,
      "step": 1233
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2834213972091675,
      "learning_rate": 0.0004178955823216767,
      "loss": 2.3933,
      "step": 1234
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.3073926568031311,
      "learning_rate": 0.00041716271502073137,
      "loss": 2.3124,
      "step": 1235
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.3054979741573334,
      "learning_rate": 0.000416430030662685,
      "loss": 2.3651,
      "step": 1236
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.295915812253952,
      "learning_rate": 0.00041569753086564173,
      "loss": 2.394,
      "step": 1237
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.3140644133090973,
      "learning_rate": 0.0004149652172472988,
      "loss": 2.3092,
      "step": 1238
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.3043701946735382,
      "learning_rate": 0.00041423309142494234,
      "loss": 2.2597,
      "step": 1239
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.3090606927871704,
      "learning_rate": 0.0004135011550154433,
      "loss": 2.4536,
      "step": 1240
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2936597466468811,
      "learning_rate": 0.0004127694096352546,
      "loss": 2.3508,
      "step": 1241
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.30298686027526855,
      "learning_rate": 0.00041203785690040743,
      "loss": 2.3052,
      "step": 1242
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.297879159450531,
      "learning_rate": 0.00041130649842650694,
      "loss": 2.3446,
      "step": 1243
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.31612586975097656,
      "learning_rate": 0.00041057533582873016,
      "loss": 2.4761,
      "step": 1244
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.3099239766597748,
      "learning_rate": 0.0004098443707218208,
      "loss": 2.2753,
      "step": 1245
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.30893808603286743,
      "learning_rate": 0.00040911360472008673,
      "loss": 2.249,
      "step": 1246
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2976721227169037,
      "learning_rate": 0.0004083830394373959,
      "loss": 2.3897,
      "step": 1247
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2993174195289612,
      "learning_rate": 0.00040765267648717324,
      "loss": 2.3635,
      "step": 1248
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.30036818981170654,
      "learning_rate": 0.00040692251748239677,
      "loss": 2.3108,
      "step": 1249
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2974975109100342,
      "learning_rate": 0.00040619256403559383,
      "loss": 2.3644,
      "step": 1250
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.31841373443603516,
      "learning_rate": 0.000405462817758838,
      "loss": 2.3926,
      "step": 1251
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2944525480270386,
      "learning_rate": 0.0004047332802637457,
      "loss": 2.2886,
      "step": 1252
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.29423993825912476,
      "learning_rate": 0.00040400395316147157,
      "loss": 2.2763,
      "step": 1253
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2978839874267578,
      "learning_rate": 0.00040327483806270627,
      "loss": 2.2569,
      "step": 1254
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2949856221675873,
      "learning_rate": 0.0004025459365776715,
      "loss": 2.3807,
      "step": 1255
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.3104931712150574,
      "learning_rate": 0.00040181725031611794,
      "loss": 2.3315,
      "step": 1256
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.29965052008628845,
      "learning_rate": 0.0004010887808873206,
      "loss": 2.348,
      "step": 1257
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.3029022812843323,
      "learning_rate": 0.00040036052990007553,
      "loss": 2.2287,
      "step": 1258
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.30464500188827515,
      "learning_rate": 0.0003996324989626967,
      "loss": 2.3622,
      "step": 1259
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.33156681060791016,
      "learning_rate": 0.00039890468968301166,
      "loss": 2.2772,
      "step": 1260
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.3079533874988556,
      "learning_rate": 0.0003981771036683591,
      "loss": 2.3667,
      "step": 1261
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.29742589592933655,
      "learning_rate": 0.00039744974252558385,
      "loss": 2.2456,
      "step": 1262
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.29985469579696655,
      "learning_rate": 0.00039672260786103463,
      "loss": 2.2636,
      "step": 1263
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2928597331047058,
      "learning_rate": 0.00039599570128055994,
      "loss": 2.2275,
      "step": 1264
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.30053603649139404,
      "learning_rate": 0.0003952690243895044,
      "loss": 2.3325,
      "step": 1265
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.28754276037216187,
      "learning_rate": 0.0003945425787927054,
      "loss": 2.3137,
      "step": 1266
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.30294978618621826,
      "learning_rate": 0.00039381636609448975,
      "loss": 2.4071,
      "step": 1267
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.308000385761261,
      "learning_rate": 0.0003930903878986693,
      "loss": 2.3332,
      "step": 1268
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.29215067625045776,
      "learning_rate": 0.00039236464580853916,
      "loss": 2.3172,
      "step": 1269
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.29460471868515015,
      "learning_rate": 0.0003916391414268718,
      "loss": 2.3419,
      "step": 1270
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.3152341842651367,
      "learning_rate": 0.00039091387635591536,
      "loss": 2.4018,
      "step": 1271
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.3063577115535736,
      "learning_rate": 0.0003901888521973894,
      "loss": 2.3629,
      "step": 1272
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.3130750358104706,
      "learning_rate": 0.0003894640705524813,
      "loss": 2.2995,
      "step": 1273
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.29029005765914917,
      "learning_rate": 0.00038873953302184284,
      "loss": 2.3227,
      "step": 1274
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.30588918924331665,
      "learning_rate": 0.000388015241205587,
      "loss": 2.3033,
      "step": 1275
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.3112208843231201,
      "learning_rate": 0.00038729119670328355,
      "loss": 2.3564,
      "step": 1276
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2908741235733032,
      "learning_rate": 0.00038656740111395665,
      "loss": 2.2704,
      "step": 1277
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.30525892972946167,
      "learning_rate": 0.00038584385603608053,
      "loss": 2.4922,
      "step": 1278
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2954728305339813,
      "learning_rate": 0.00038512056306757615,
      "loss": 2.3605,
      "step": 1279
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.31617048382759094,
      "learning_rate": 0.0003843975238058075,
      "loss": 2.3901,
      "step": 1280
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.300578236579895,
      "learning_rate": 0.00038367473984757863,
      "loss": 2.2619,
      "step": 1281
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2949879765510559,
      "learning_rate": 0.0003829522127891296,
      "loss": 2.2539,
      "step": 1282
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.29724204540252686,
      "learning_rate": 0.0003822299442261329,
      "loss": 2.3681,
      "step": 1283
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.31423670053482056,
      "learning_rate": 0.00038150793575369063,
      "loss": 2.2977,
      "step": 1284
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.30838772654533386,
      "learning_rate": 0.0003807861889663299,
      "loss": 2.2302,
      "step": 1285
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.3131427764892578,
      "learning_rate": 0.0003800647054580006,
      "loss": 2.3574,
      "step": 1286
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.31080830097198486,
      "learning_rate": 0.00037934348682207064,
      "loss": 2.2939,
      "step": 1287
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.31118300557136536,
      "learning_rate": 0.00037862253465132306,
      "loss": 2.4387,
      "step": 1288
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.30553245544433594,
      "learning_rate": 0.00037790185053795245,
      "loss": 2.2424,
      "step": 1289
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.3077365458011627,
      "learning_rate": 0.0003771814360735616,
      "loss": 2.2344,
      "step": 1290
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.29994872212409973,
      "learning_rate": 0.00037646129284915755,
      "loss": 2.3913,
      "step": 1291
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.30683764815330505,
      "learning_rate": 0.00037574142245514825,
      "loss": 2.3184,
      "step": 1292
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.3008595108985901,
      "learning_rate": 0.0003750218264813393,
      "loss": 2.369,
      "step": 1293
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.28859227895736694,
      "learning_rate": 0.0003743025065169305,
      "loss": 2.1966,
      "step": 1294
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.3034055233001709,
      "learning_rate": 0.0003735834641505116,
      "loss": 2.3802,
      "step": 1295
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.291421502828598,
      "learning_rate": 0.00037286470097005954,
      "loss": 2.2653,
      "step": 1296
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.30840426683425903,
      "learning_rate": 0.0003721462185629347,
      "loss": 2.2525,
      "step": 1297
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.30564290285110474,
      "learning_rate": 0.00037142801851587707,
      "loss": 2.3723,
      "step": 1298
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3013574182987213,
      "learning_rate": 0.00037071010241500357,
      "loss": 2.2986,
      "step": 1299
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3126370906829834,
      "learning_rate": 0.00036999247184580383,
      "loss": 2.2935,
      "step": 1300
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.321714848279953,
      "learning_rate": 0.00036927512839313636,
      "loss": 2.3748,
      "step": 1301
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.29786285758018494,
      "learning_rate": 0.0003685580736412268,
      "loss": 2.3242,
      "step": 1302
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3033115863800049,
      "learning_rate": 0.000367841309173662,
      "loss": 2.2898,
      "step": 1303
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.325803279876709,
      "learning_rate": 0.0003671248365733883,
      "loss": 2.4144,
      "step": 1304
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.30198052525520325,
      "learning_rate": 0.0003664086574227075,
      "loss": 2.2655,
      "step": 1305
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2819981276988983,
      "learning_rate": 0.000365692773303273,
      "loss": 2.3179,
      "step": 1306
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.29360446333885193,
      "learning_rate": 0.00036497718579608696,
      "loss": 2.4101,
      "step": 1307
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2970242500305176,
      "learning_rate": 0.0003642618964814964,
      "loss": 2.3202,
      "step": 1308
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.28141728043556213,
      "learning_rate": 0.00036354690693918946,
      "loss": 2.3522,
      "step": 1309
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.295284241437912,
      "learning_rate": 0.00036283221874819284,
      "loss": 2.2804,
      "step": 1310
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2915458679199219,
      "learning_rate": 0.0003621178334868672,
      "loss": 2.3464,
      "step": 1311
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3049074411392212,
      "learning_rate": 0.00036140375273290476,
      "loss": 2.3927,
      "step": 1312
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2985147535800934,
      "learning_rate": 0.0003606899780633245,
      "loss": 2.3435,
      "step": 1313
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.29284408688545227,
      "learning_rate": 0.0003599765110544699,
      "loss": 2.3428,
      "step": 1314
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.30021822452545166,
      "learning_rate": 0.0003592633532820052,
      "loss": 2.3273,
      "step": 1315
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3141315281391144,
      "learning_rate": 0.0003585505063209109,
      "loss": 2.3268,
      "step": 1316
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3155769407749176,
      "learning_rate": 0.00035783797174548194,
      "loss": 2.4102,
      "step": 1317
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.32073161005973816,
      "learning_rate": 0.00035712575112932277,
      "loss": 2.3896,
      "step": 1318
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.31024593114852905,
      "learning_rate": 0.000356413846045345,
      "loss": 2.2626,
      "step": 1319
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.3001432418823242,
      "learning_rate": 0.000355702258065763,
      "loss": 2.2745,
      "step": 1320
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.31484687328338623,
      "learning_rate": 0.0003549909887620909,
      "loss": 2.3263,
      "step": 1321
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.30440041422843933,
      "learning_rate": 0.00035428003970513914,
      "loss": 2.2428,
      "step": 1322
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.30284592509269714,
      "learning_rate": 0.00035356941246501085,
      "loss": 2.3414,
      "step": 1323
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.30005958676338196,
      "learning_rate": 0.0003528591086110984,
      "loss": 2.2536,
      "step": 1324
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2938177287578583,
      "learning_rate": 0.00035214912971208,
      "loss": 2.3496,
      "step": 1325
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2921377718448639,
      "learning_rate": 0.0003514394773359163,
      "loss": 2.3486,
      "step": 1326
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2955338656902313,
      "learning_rate": 0.0003507301530498469,
      "loss": 2.4264,
      "step": 1327
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.3006444275379181,
      "learning_rate": 0.00035002115842038646,
      "loss": 2.3423,
      "step": 1328
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.29187139868736267,
      "learning_rate": 0.00034931249501332195,
      "loss": 2.3392,
      "step": 1329
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.3205322027206421,
      "learning_rate": 0.00034860416439370885,
      "loss": 2.3607,
      "step": 1330
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2974936068058014,
      "learning_rate": 0.0003478961681258674,
      "loss": 2.3622,
      "step": 1331
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.3131506145000458,
      "learning_rate": 0.0003471885077733796,
      "loss": 2.2776,
      "step": 1332
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.31280574202537537,
      "learning_rate": 0.0003464811848990859,
      "loss": 2.359,
      "step": 1333
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.4655206799507141,
      "learning_rate": 0.00034577420106508063,
      "loss": 2.3289,
      "step": 1334
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.291390985250473,
      "learning_rate": 0.0003450675578327105,
      "loss": 2.2618,
      "step": 1335
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.3067641258239746,
      "learning_rate": 0.000344361256762569,
      "loss": 2.3571,
      "step": 1336
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.3313268721103668,
      "learning_rate": 0.00034365529941449456,
      "loss": 2.2526,
      "step": 1337
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.3082340657711029,
      "learning_rate": 0.0003429496873475664,
      "loss": 2.303,
      "step": 1338
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.3109550476074219,
      "learning_rate": 0.0003422444221201009,
      "loss": 2.3095,
      "step": 1339
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.318985253572464,
      "learning_rate": 0.0003415395052896487,
      "loss": 2.3112,
      "step": 1340
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.30483776330947876,
      "learning_rate": 0.0003408349384129912,
      "loss": 2.2475,
      "step": 1341
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.3064630925655365,
      "learning_rate": 0.00034013072304613643,
      "loss": 2.3081,
      "step": 1342
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.30003559589385986,
      "learning_rate": 0.00033942686074431674,
      "loss": 2.2871,
      "step": 1343
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.3283809423446655,
      "learning_rate": 0.0003387233530619843,
      "loss": 2.3794,
      "step": 1344
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2906898558139801,
      "learning_rate": 0.0003380202015528084,
      "loss": 2.3243,
      "step": 1345
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.3117261528968811,
      "learning_rate": 0.0003373174077696715,
      "loss": 2.3032,
      "step": 1346
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2920597195625305,
      "learning_rate": 0.0003366149732646661,
      "loss": 2.2625,
      "step": 1347
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2892654240131378,
      "learning_rate": 0.00033591289958909143,
      "loss": 2.3086,
      "step": 1348
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.3019716143608093,
      "learning_rate": 0.00033521118829344954,
      "loss": 2.2527,
      "step": 1349
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2887209951877594,
      "learning_rate": 0.0003345098409274423,
      "loss": 2.3143,
      "step": 1350
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.30358442664146423,
      "learning_rate": 0.00033380885903996796,
      "loss": 2.3757,
      "step": 1351
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.30145829916000366,
      "learning_rate": 0.00033310824417911766,
      "loss": 2.3652,
      "step": 1352
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2988254129886627,
      "learning_rate": 0.00033240799789217184,
      "loss": 2.2776,
      "step": 1353
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2926655113697052,
      "learning_rate": 0.00033170812172559694,
      "loss": 2.3036,
      "step": 1354
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.31377482414245605,
      "learning_rate": 0.000331008617225042,
      "loss": 2.3598,
      "step": 1355
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2977752983570099,
      "learning_rate": 0.0003303094859353355,
      "loss": 2.4127,
      "step": 1356
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.320299357175827,
      "learning_rate": 0.0003296107294004812,
      "loss": 2.2521,
      "step": 1357
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.3052029609680176,
      "learning_rate": 0.0003289123491636559,
      "loss": 2.2656,
      "step": 1358
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.29997339844703674,
      "learning_rate": 0.00032821434676720443,
      "loss": 2.3232,
      "step": 1359
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.29920655488967896,
      "learning_rate": 0.00032751672375263836,
      "loss": 2.2727,
      "step": 1360
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.29876425862312317,
      "learning_rate": 0.0003268194816606305,
      "loss": 2.3136,
      "step": 1361
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.30789604783058167,
      "learning_rate": 0.00032612262203101267,
      "loss": 2.3372,
      "step": 1362
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3064424395561218,
      "learning_rate": 0.00032542614640277225,
      "loss": 2.3836,
      "step": 1363
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3101542294025421,
      "learning_rate": 0.0003247300563140481,
      "loss": 2.3316,
      "step": 1364
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2897811830043793,
      "learning_rate": 0.00032403435330212807,
      "loss": 2.2693,
      "step": 1365
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.30199337005615234,
      "learning_rate": 0.00032333903890344515,
      "loss": 2.4864,
      "step": 1366
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3016076385974884,
      "learning_rate": 0.00032264411465357333,
      "loss": 2.3429,
      "step": 1367
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3076232373714447,
      "learning_rate": 0.00032194958208722654,
      "loss": 2.3325,
      "step": 1368
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3082130551338196,
      "learning_rate": 0.00032125544273825204,
      "loss": 2.2686,
      "step": 1369
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3023010790348053,
      "learning_rate": 0.0003205616981396297,
      "loss": 2.2709,
      "step": 1370
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.29259955883026123,
      "learning_rate": 0.00031986834982346713,
      "loss": 2.2033,
      "step": 1371
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3060935437679291,
      "learning_rate": 0.00031917539932099694,
      "loss": 2.3687,
      "step": 1372
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.31505832076072693,
      "learning_rate": 0.00031848284816257336,
      "loss": 2.361,
      "step": 1373
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3082973062992096,
      "learning_rate": 0.0003177906978776682,
      "loss": 2.2898,
      "step": 1374
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.29444223642349243,
      "learning_rate": 0.0003170989499948683,
      "loss": 2.2845,
      "step": 1375
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3010799288749695,
      "learning_rate": 0.0003164076060418719,
      "loss": 2.3604,
      "step": 1376
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2844693958759308,
      "learning_rate": 0.000315716667545485,
      "loss": 2.3597,
      "step": 1377
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3119623064994812,
      "learning_rate": 0.00031502613603161836,
      "loss": 2.251,
      "step": 1378
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2973722815513611,
      "learning_rate": 0.00031433601302528335,
      "loss": 2.3526,
      "step": 1379
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3073907196521759,
      "learning_rate": 0.00031364630005058995,
      "loss": 2.3819,
      "step": 1380
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2911858856678009,
      "learning_rate": 0.0003129569986307422,
      "loss": 2.3507,
      "step": 1381
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.30882731080055237,
      "learning_rate": 0.00031226811028803515,
      "loss": 2.3851,
      "step": 1382
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2909899055957794,
      "learning_rate": 0.00031157963654385173,
      "loss": 2.2679,
      "step": 1383
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2907232940196991,
      "learning_rate": 0.0003108915789186592,
      "loss": 2.327,
      "step": 1384
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2924480140209198,
      "learning_rate": 0.00031020393893200604,
      "loss": 2.2133,
      "step": 1385
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2933991551399231,
      "learning_rate": 0.00030951671810251823,
      "loss": 2.2992,
      "step": 1386
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3049492835998535,
      "learning_rate": 0.0003088299179478959,
      "loss": 2.4203,
      "step": 1387
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.31511014699935913,
      "learning_rate": 0.0003081435399849104,
      "loss": 2.31,
      "step": 1388
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3015027642250061,
      "learning_rate": 0.0003074575857294004,
      "loss": 2.2877,
      "step": 1389
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3048892915248871,
      "learning_rate": 0.0003067720566962691,
      "loss": 2.3984,
      "step": 1390
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3025968372821808,
      "learning_rate": 0.0003060869543994806,
      "loss": 2.3835,
      "step": 1391
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3093753755092621,
      "learning_rate": 0.0003054022803520562,
      "loss": 2.3245,
      "step": 1392
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2907174825668335,
      "learning_rate": 0.0003047180360660721,
      "loss": 2.3274,
      "step": 1393
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3124445378780365,
      "learning_rate": 0.00030403422305265475,
      "loss": 2.3117,
      "step": 1394
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.29822486639022827,
      "learning_rate": 0.0003033508428219785,
      "loss": 2.3975,
      "step": 1395
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.32173246145248413,
      "learning_rate": 0.00030266789688326184,
      "loss": 2.2701,
      "step": 1396
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.30054742097854614,
      "learning_rate": 0.00030198538674476393,
      "loss": 2.3612,
      "step": 1397
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.32206469774246216,
      "learning_rate": 0.00030130331391378185,
      "loss": 2.3234,
      "step": 1398
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.31389135122299194,
      "learning_rate": 0.0003006216798966468,
      "loss": 2.3137,
      "step": 1399
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.28286856412887573,
      "learning_rate": 0.00029994048619872034,
      "loss": 2.3229,
      "step": 1400
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.30735406279563904,
      "learning_rate": 0.0002992597343243927,
      "loss": 2.22,
      "step": 1401
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3016727864742279,
      "learning_rate": 0.0002985794257770773,
      "loss": 2.3315,
      "step": 1402
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3255265951156616,
      "learning_rate": 0.0002978995620592092,
      "loss": 2.2978,
      "step": 1403
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3089118003845215,
      "learning_rate": 0.0002972201446722405,
      "loss": 2.3247,
      "step": 1404
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3023015558719635,
      "learning_rate": 0.00029654117511663803,
      "loss": 2.2073,
      "step": 1405
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3259861469268799,
      "learning_rate": 0.0002958626548918795,
      "loss": 2.3633,
      "step": 1406
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.33771592378616333,
      "learning_rate": 0.00029518458549645014,
      "loss": 2.3263,
      "step": 1407
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.30650609731674194,
      "learning_rate": 0.00029450696842783954,
      "loss": 2.292,
      "step": 1408
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.3169569969177246,
      "learning_rate": 0.00029382980518253865,
      "loss": 2.2742,
      "step": 1409
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.3017321527004242,
      "learning_rate": 0.00029315309725603595,
      "loss": 2.2406,
      "step": 1410
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.3039901852607727,
      "learning_rate": 0.00029247684614281446,
      "loss": 2.3711,
      "step": 1411
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.29898086190223694,
      "learning_rate": 0.0002918010533363481,
      "loss": 2.3241,
      "step": 1412
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.32402849197387695,
      "learning_rate": 0.0002911257203290987,
      "loss": 2.2513,
      "step": 1413
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.3165958523750305,
      "learning_rate": 0.00029045084861251314,
      "loss": 2.4017,
      "step": 1414
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.31164729595184326,
      "learning_rate": 0.00028977643967701897,
      "loss": 2.3959,
      "step": 1415
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.3008746802806854,
      "learning_rate": 0.00028910249501202156,
      "loss": 2.2349,
      "step": 1416
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.29877233505249023,
      "learning_rate": 0.00028842901610590165,
      "loss": 2.3344,
      "step": 1417
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.29289352893829346,
      "learning_rate": 0.00028775600444601123,
      "loss": 2.3134,
      "step": 1418
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2805945873260498,
      "learning_rate": 0.00028708346151866973,
      "loss": 2.2315,
      "step": 1419
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2982195317745209,
      "learning_rate": 0.0002864113888091622,
      "loss": 2.3143,
      "step": 1420
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.29399919509887695,
      "learning_rate": 0.0002857397878017348,
      "loss": 2.2899,
      "step": 1421
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.28516048192977905,
      "learning_rate": 0.00028506865997959173,
      "loss": 2.4249,
      "step": 1422
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.301901638507843,
      "learning_rate": 0.000284398006824893,
      "loss": 2.3788,
      "step": 1423
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.312300443649292,
      "learning_rate": 0.00028372782981874963,
      "loss": 2.2406,
      "step": 1424
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.3154506981372833,
      "learning_rate": 0.00028305813044122096,
      "loss": 2.3553,
      "step": 1425
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.320153146982193,
      "learning_rate": 0.0002823889101713122,
      "loss": 2.3559,
      "step": 1426
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.3173193335533142,
      "learning_rate": 0.0002817201704869701,
      "loss": 2.212,
      "step": 1427
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.30394861102104187,
      "learning_rate": 0.00028105191286508,
      "loss": 2.2861,
      "step": 1428
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.3077945411205292,
      "learning_rate": 0.00028038413878146245,
      "loss": 2.3031,
      "step": 1429
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2982640266418457,
      "learning_rate": 0.00027971684971087073,
      "loss": 2.2393,
      "step": 1430
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.3034318685531616,
      "learning_rate": 0.00027905004712698643,
      "loss": 2.287,
      "step": 1431
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.30359724164009094,
      "learning_rate": 0.0002783837325024167,
      "loss": 2.1983,
      "step": 1432
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.30559882521629333,
      "learning_rate": 0.00027771790730869153,
      "loss": 2.2705,
      "step": 1433
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.30240869522094727,
      "learning_rate": 0.0002770525730162599,
      "loss": 2.2973,
      "step": 1434
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2990531921386719,
      "learning_rate": 0.00027638773109448645,
      "loss": 2.2612,
      "step": 1435
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2906437814235687,
      "learning_rate": 0.00027572338301164824,
      "loss": 2.3362,
      "step": 1436
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.295552521944046,
      "learning_rate": 0.0002750595302349324,
      "loss": 2.3854,
      "step": 1437
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.30111783742904663,
      "learning_rate": 0.00027439617423043145,
      "loss": 2.2679,
      "step": 1438
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2993066608905792,
      "learning_rate": 0.00027373331646314114,
      "loss": 2.3269,
      "step": 1439
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.30300527811050415,
      "learning_rate": 0.0002730709583969572,
      "loss": 2.3059,
      "step": 1440
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2830442488193512,
      "learning_rate": 0.0002724091014946711,
      "loss": 2.3844,
      "step": 1441
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.3188560903072357,
      "learning_rate": 0.00027174774721796824,
      "loss": 2.3125,
      "step": 1442
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.30848172307014465,
      "learning_rate": 0.0002710868970274232,
      "loss": 2.2886,
      "step": 1443
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2866440415382385,
      "learning_rate": 0.0002704265523824982,
      "loss": 2.3166,
      "step": 1444
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.30151671171188354,
      "learning_rate": 0.00026976671474153826,
      "loss": 2.2659,
      "step": 1445
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.3057786822319031,
      "learning_rate": 0.00026910738556176886,
      "loss": 2.377,
      "step": 1446
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.29115399718284607,
      "learning_rate": 0.0002684485662992929,
      "loss": 2.3006,
      "step": 1447
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2999427914619446,
      "learning_rate": 0.0002677902584090869,
      "loss": 2.3131,
      "step": 1448
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.31151387095451355,
      "learning_rate": 0.00026713246334499774,
      "loss": 2.3501,
      "step": 1449
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2904815077781677,
      "learning_rate": 0.00026647518255974023,
      "loss": 2.387,
      "step": 1450
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.29715290665626526,
      "learning_rate": 0.0002658184175048934,
      "loss": 2.3414,
      "step": 1451
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.29388681054115295,
      "learning_rate": 0.00026516216963089694,
      "loss": 2.3389,
      "step": 1452
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.30883729457855225,
      "learning_rate": 0.0002645064403870488,
      "loss": 2.3635,
      "step": 1453
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.30116257071495056,
      "learning_rate": 0.0002638512312215011,
      "loss": 2.3328,
      "step": 1454
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3066740930080414,
      "learning_rate": 0.0002631965435812575,
      "loss": 2.2785,
      "step": 1455
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.30803778767585754,
      "learning_rate": 0.00026254237891217046,
      "loss": 2.3444,
      "step": 1456
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.30522942543029785,
      "learning_rate": 0.0002618887386589367,
      "loss": 2.2614,
      "step": 1457
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3054317533969879,
      "learning_rate": 0.0002612356242650949,
      "loss": 2.2999,
      "step": 1458
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3200596868991852,
      "learning_rate": 0.0002605830371730229,
      "loss": 2.2929,
      "step": 1459
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3089091181755066,
      "learning_rate": 0.0002599309788239339,
      "loss": 2.3614,
      "step": 1460
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3071094751358032,
      "learning_rate": 0.00025927945065787306,
      "loss": 2.3025,
      "step": 1461
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3056503236293793,
      "learning_rate": 0.0002586284541137145,
      "loss": 2.2644,
      "step": 1462
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3081713318824768,
      "learning_rate": 0.00025797799062915905,
      "loss": 2.2975,
      "step": 1463
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3225601017475128,
      "learning_rate": 0.00025732806164072966,
      "loss": 2.3053,
      "step": 1464
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.31230485439300537,
      "learning_rate": 0.00025667866858376874,
      "loss": 2.2972,
      "step": 1465
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.32091447710990906,
      "learning_rate": 0.0002560298128924358,
      "loss": 2.2982,
      "step": 1466
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.32540589570999146,
      "learning_rate": 0.0002553814959997032,
      "loss": 2.2634,
      "step": 1467
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.32578158378601074,
      "learning_rate": 0.00025473371933735334,
      "loss": 2.3698,
      "step": 1468
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.31097403168678284,
      "learning_rate": 0.00025408648433597534,
      "loss": 2.3132,
      "step": 1469
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.297000527381897,
      "learning_rate": 0.00025343979242496283,
      "loss": 2.2881,
      "step": 1470
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.296414852142334,
      "learning_rate": 0.00025279364503250925,
      "loss": 2.3758,
      "step": 1471
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2941439747810364,
      "learning_rate": 0.0002521480435856056,
      "loss": 2.3197,
      "step": 1472
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.31058958172798157,
      "learning_rate": 0.0002515029895100378,
      "loss": 2.43,
      "step": 1473
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2842855453491211,
      "learning_rate": 0.0002508584842303822,
      "loss": 2.2841,
      "step": 1474
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2976178228855133,
      "learning_rate": 0.0002502145291700038,
      "loss": 2.3438,
      "step": 1475
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3020893633365631,
      "learning_rate": 0.0002495711257510517,
      "loss": 2.2201,
      "step": 1476
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.30278995633125305,
      "learning_rate": 0.0002489282753944575,
      "loss": 2.3686,
      "step": 1477
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2905269265174866,
      "learning_rate": 0.00024828597951993093,
      "loss": 2.3111,
      "step": 1478
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3038938343524933,
      "learning_rate": 0.00024764423954595706,
      "loss": 2.3449,
      "step": 1479
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2958216965198517,
      "learning_rate": 0.0002470030568897938,
      "loss": 2.1888,
      "step": 1480
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.30887433886528015,
      "learning_rate": 0.00024636243296746773,
      "loss": 2.2245,
      "step": 1481
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.30145496129989624,
      "learning_rate": 0.0002457223691937716,
      "loss": 2.1972,
      "step": 1482
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3131498396396637,
      "learning_rate": 0.0002450828669822613,
      "loss": 2.3281,
      "step": 1483
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3062089681625366,
      "learning_rate": 0.00024444392774525253,
      "loss": 2.2978,
      "step": 1484
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2997904121875763,
      "learning_rate": 0.00024380555289381733,
      "loss": 2.2062,
      "step": 1485
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.31126463413238525,
      "learning_rate": 0.00024316774383778184,
      "loss": 2.4264,
      "step": 1486
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3077186644077301,
      "learning_rate": 0.0002425305019857222,
      "loss": 2.2781,
      "step": 1487
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.31103384494781494,
      "learning_rate": 0.00024189382874496184,
      "loss": 2.3614,
      "step": 1488
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3027143180370331,
      "learning_rate": 0.00024125772552156916,
      "loss": 2.3916,
      "step": 1489
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.31482428312301636,
      "learning_rate": 0.00024062219372035292,
      "loss": 2.2587,
      "step": 1490
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.30958226323127747,
      "learning_rate": 0.00023998723474486007,
      "loss": 2.2886,
      "step": 1491
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2949078381061554,
      "learning_rate": 0.00023935284999737272,
      "loss": 2.2826,
      "step": 1492
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2989675998687744,
      "learning_rate": 0.00023871904087890505,
      "loss": 2.3712,
      "step": 1493
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.30461686849594116,
      "learning_rate": 0.00023808580878919945,
      "loss": 2.2456,
      "step": 1494
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.3056599795818329,
      "learning_rate": 0.00023745315512672398,
      "loss": 2.3389,
      "step": 1495
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.32317715883255005,
      "learning_rate": 0.0002368210812886698,
      "loss": 2.3178,
      "step": 1496
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.30584776401519775,
      "learning_rate": 0.0002361895886709471,
      "loss": 2.1907,
      "step": 1497
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.3289037048816681,
      "learning_rate": 0.0002355586786681823,
      "loss": 2.2641,
      "step": 1498
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.3318312466144562,
      "learning_rate": 0.00023492835267371575,
      "loss": 2.2819,
      "step": 1499
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.3129958212375641,
      "learning_rate": 0.0002342986120795978,
      "loss": 2.2841,
      "step": 1500
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.3206716477870941,
      "learning_rate": 0.0002336694582765857,
      "loss": 2.2638,
      "step": 1501
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.33235445618629456,
      "learning_rate": 0.00023304089265414085,
      "loss": 2.3245,
      "step": 1502
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.3120078444480896,
      "learning_rate": 0.00023241291660042613,
      "loss": 2.2754,
      "step": 1503
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.28912389278411865,
      "learning_rate": 0.0002317855315023018,
      "loss": 2.2453,
      "step": 1504
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.30557698011398315,
      "learning_rate": 0.00023115873874532324,
      "loss": 2.3957,
      "step": 1505
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.29483115673065186,
      "learning_rate": 0.00023053253971373796,
      "loss": 2.3319,
      "step": 1506
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2966075837612152,
      "learning_rate": 0.00022990693579048166,
      "loss": 2.2668,
      "step": 1507
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2878841161727905,
      "learning_rate": 0.00022928192835717644,
      "loss": 2.2859,
      "step": 1508
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.28425660729408264,
      "learning_rate": 0.00022865751879412634,
      "loss": 2.3701,
      "step": 1509
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.311758428812027,
      "learning_rate": 0.00022803370848031585,
      "loss": 2.3032,
      "step": 1510
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.29259946942329407,
      "learning_rate": 0.00022741049879340542,
      "loss": 2.3257,
      "step": 1511
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2858245372772217,
      "learning_rate": 0.00022678789110972897,
      "loss": 2.3452,
      "step": 1512
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.30386343598365784,
      "learning_rate": 0.00022616588680429155,
      "loss": 2.3362,
      "step": 1513
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2939131259918213,
      "learning_rate": 0.00022554448725076526,
      "loss": 2.1744,
      "step": 1514
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2945854663848877,
      "learning_rate": 0.0002249236938214863,
      "loss": 2.2482,
      "step": 1515
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3378181457519531,
      "learning_rate": 0.00022430350788745296,
      "loss": 2.3801,
      "step": 1516
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3064638078212738,
      "learning_rate": 0.00022368393081832166,
      "loss": 2.3186,
      "step": 1517
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3150656521320343,
      "learning_rate": 0.00022306496398240383,
      "loss": 2.2709,
      "step": 1518
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3291935324668884,
      "learning_rate": 0.00022244660874666373,
      "loss": 2.2468,
      "step": 1519
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3186374604701996,
      "learning_rate": 0.00022182886647671452,
      "loss": 2.3781,
      "step": 1520
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3218635320663452,
      "learning_rate": 0.0002212117385368157,
      "loss": 2.285,
      "step": 1521
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3111070692539215,
      "learning_rate": 0.00022059522628987038,
      "loss": 2.2829,
      "step": 1522
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.30482104420661926,
      "learning_rate": 0.00021997933109742162,
      "loss": 2.2685,
      "step": 1523
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.35895174741744995,
      "learning_rate": 0.00021936405431964969,
      "loss": 2.3897,
      "step": 1524
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3036831319332123,
      "learning_rate": 0.00021874939731536926,
      "loss": 2.2704,
      "step": 1525
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3020024597644806,
      "learning_rate": 0.00021813536144202656,
      "loss": 2.2897,
      "step": 1526
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3045869469642639,
      "learning_rate": 0.00021752194805569553,
      "loss": 2.1645,
      "step": 1527
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.30664771795272827,
      "learning_rate": 0.0002169091585110754,
      "loss": 2.2351,
      "step": 1528
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.2924720048904419,
      "learning_rate": 0.00021629699416148828,
      "loss": 2.3376,
      "step": 1529
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3068137466907501,
      "learning_rate": 0.000215685456358875,
      "loss": 2.2905,
      "step": 1530
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.301412433385849,
      "learning_rate": 0.00021507454645379258,
      "loss": 2.2892,
      "step": 1531
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3081165552139282,
      "learning_rate": 0.00021446426579541184,
      "loss": 2.4213,
      "step": 1532
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3051658272743225,
      "learning_rate": 0.00021385461573151387,
      "loss": 2.3394,
      "step": 1533
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3154163360595703,
      "learning_rate": 0.00021324559760848677,
      "loss": 2.3937,
      "step": 1534
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.2916833162307739,
      "learning_rate": 0.00021263721277132303,
      "loss": 2.35,
      "step": 1535
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3106239140033722,
      "learning_rate": 0.0002120294625636171,
      "loss": 2.2843,
      "step": 1536
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3133402466773987,
      "learning_rate": 0.0002114223483275613,
      "loss": 2.125,
      "step": 1537
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2996361255645752,
      "learning_rate": 0.0002108158714039435,
      "loss": 2.292,
      "step": 1538
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.3095130920410156,
      "learning_rate": 0.00021021003313214455,
      "loss": 2.3483,
      "step": 1539
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.30220115184783936,
      "learning_rate": 0.00020960483485013432,
      "loss": 2.2855,
      "step": 1540
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.29728519916534424,
      "learning_rate": 0.0002090002778944694,
      "loss": 2.2991,
      "step": 1541
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.31422939896583557,
      "learning_rate": 0.00020839636360029025,
      "loss": 2.279,
      "step": 1542
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.31080710887908936,
      "learning_rate": 0.00020779309330131818,
      "loss": 2.349,
      "step": 1543
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.31519243121147156,
      "learning_rate": 0.00020719046832985184,
      "loss": 2.3025,
      "step": 1544
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.3388166129589081,
      "learning_rate": 0.0002065884900167646,
      "loss": 2.4063,
      "step": 1545
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2962445318698883,
      "learning_rate": 0.0002059871596915024,
      "loss": 2.2494,
      "step": 1546
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.31322237849235535,
      "learning_rate": 0.0002053864786820795,
      "loss": 2.318,
      "step": 1547
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.3007246255874634,
      "learning_rate": 0.00020478644831507627,
      "loss": 2.3197,
      "step": 1548
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.31533828377723694,
      "learning_rate": 0.00020418706991563634,
      "loss": 2.3116,
      "step": 1549
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.30647069215774536,
      "learning_rate": 0.00020358834480746363,
      "loss": 2.3481,
      "step": 1550
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.31025445461273193,
      "learning_rate": 0.0002029902743128188,
      "loss": 2.4022,
      "step": 1551
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.30067211389541626,
      "learning_rate": 0.0002023928597525174,
      "loss": 2.3306,
      "step": 1552
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2997124195098877,
      "learning_rate": 0.00020179610244592595,
      "loss": 2.2779,
      "step": 1553
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.31288328766822815,
      "learning_rate": 0.00020120000371095937,
      "loss": 2.3722,
      "step": 1554
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.32435736060142517,
      "learning_rate": 0.0002006045648640787,
      "loss": 2.3272,
      "step": 1555
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2890874147415161,
      "learning_rate": 0.00020000978722028713,
      "loss": 2.2744,
      "step": 1556
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.29660969972610474,
      "learning_rate": 0.00019941567209312767,
      "loss": 2.2713,
      "step": 1557
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.30767756700515747,
      "learning_rate": 0.00019882222079468036,
      "loss": 2.3243,
      "step": 1558
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.30329757928848267,
      "learning_rate": 0.0001982294346355595,
      "loss": 2.2199,
      "step": 1559
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.30780476331710815,
      "learning_rate": 0.00019763731492490976,
      "loss": 2.3758,
      "step": 1560
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.28758370876312256,
      "learning_rate": 0.00019704586297040422,
      "loss": 2.2659,
      "step": 1561
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.314968079328537,
      "learning_rate": 0.0001964550800782417,
      "loss": 2.2942,
      "step": 1562
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3048268258571625,
      "learning_rate": 0.00019586496755314288,
      "loss": 2.3316,
      "step": 1563
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.31245580315589905,
      "learning_rate": 0.00019527552669834798,
      "loss": 2.2913,
      "step": 1564
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.28788191080093384,
      "learning_rate": 0.0001946867588156142,
      "loss": 2.3207,
      "step": 1565
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.31263014674186707,
      "learning_rate": 0.00019409866520521258,
      "loss": 2.3111,
      "step": 1566
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.30504223704338074,
      "learning_rate": 0.00019351124716592455,
      "loss": 2.1979,
      "step": 1567
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.31201839447021484,
      "learning_rate": 0.0001929245059950397,
      "loss": 2.3388,
      "step": 1568
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.29986080527305603,
      "learning_rate": 0.0001923384429883533,
      "loss": 2.3695,
      "step": 1569
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.31516173481941223,
      "learning_rate": 0.00019175305944016237,
      "loss": 2.2604,
      "step": 1570
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3141469657421112,
      "learning_rate": 0.00019116835664326326,
      "loss": 2.33,
      "step": 1571
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3264119327068329,
      "learning_rate": 0.0001905843358889497,
      "loss": 2.3098,
      "step": 1572
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.2967489957809448,
      "learning_rate": 0.00019000099846700836,
      "loss": 2.279,
      "step": 1573
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3047792315483093,
      "learning_rate": 0.00018941834566571692,
      "loss": 2.2436,
      "step": 1574
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3284531831741333,
      "learning_rate": 0.00018883637877184145,
      "loss": 2.3391,
      "step": 1575
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3252377510070801,
      "learning_rate": 0.00018825509907063325,
      "loss": 2.3893,
      "step": 1576
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.32534918189048767,
      "learning_rate": 0.00018767450784582557,
      "loss": 2.4066,
      "step": 1577
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3274228572845459,
      "learning_rate": 0.00018709460637963122,
      "loss": 2.3589,
      "step": 1578
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.30476027727127075,
      "learning_rate": 0.00018651539595274013,
      "loss": 2.2511,
      "step": 1579
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.31444448232650757,
      "learning_rate": 0.00018593687784431578,
      "loss": 2.3429,
      "step": 1580
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.32797104120254517,
      "learning_rate": 0.00018535905333199248,
      "loss": 2.2815,
      "step": 1581
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3183436691761017,
      "learning_rate": 0.0001847819236918733,
      "loss": 2.2786,
      "step": 1582
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.29540902376174927,
      "learning_rate": 0.00018420549019852655,
      "loss": 2.2119,
      "step": 1583
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2922125458717346,
      "learning_rate": 0.00018362975412498266,
      "loss": 2.2811,
      "step": 1584
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.312339186668396,
      "learning_rate": 0.00018305471674273261,
      "loss": 2.278,
      "step": 1585
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.29891425371170044,
      "learning_rate": 0.0001824803793217237,
      "loss": 2.4121,
      "step": 1586
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3269616365432739,
      "learning_rate": 0.00018190674313035737,
      "loss": 2.2634,
      "step": 1587
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3023204505443573,
      "learning_rate": 0.00018133380943548716,
      "loss": 2.3176,
      "step": 1588
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3088812530040741,
      "learning_rate": 0.00018076157950241452,
      "loss": 2.3582,
      "step": 1589
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.30174145102500916,
      "learning_rate": 0.00018019005459488652,
      "loss": 2.2756,
      "step": 1590
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3169102370738983,
      "learning_rate": 0.00017961923597509388,
      "loss": 2.1899,
      "step": 1591
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2974310517311096,
      "learning_rate": 0.00017904912490366722,
      "loss": 2.2987,
      "step": 1592
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3062911033630371,
      "learning_rate": 0.00017847972263967433,
      "loss": 2.3125,
      "step": 1593
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3036291301250458,
      "learning_rate": 0.0001779110304406177,
      "loss": 2.2585,
      "step": 1594
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3240666687488556,
      "learning_rate": 0.000177343049562432,
      "loss": 2.2696,
      "step": 1595
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3093837797641754,
      "learning_rate": 0.0001767757812594807,
      "loss": 2.3405,
      "step": 1596
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.31116706132888794,
      "learning_rate": 0.0001762092267845534,
      "loss": 2.3459,
      "step": 1597
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.33090460300445557,
      "learning_rate": 0.00017564338738886365,
      "loss": 2.3978,
      "step": 1598
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.30900201201438904,
      "learning_rate": 0.0001750782643220457,
      "loss": 2.3088,
      "step": 1599
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3091129958629608,
      "learning_rate": 0.00017451385883215166,
      "loss": 2.3907,
      "step": 1600
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.29903385043144226,
      "learning_rate": 0.00017395017216564863,
      "loss": 2.2796,
      "step": 1601
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3248565196990967,
      "learning_rate": 0.00017338720556741687,
      "loss": 2.2601,
      "step": 1602
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2984296381473541,
      "learning_rate": 0.00017282496028074606,
      "loss": 2.2494,
      "step": 1603
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3206033706665039,
      "learning_rate": 0.00017226343754733254,
      "loss": 2.2865,
      "step": 1604
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3142213523387909,
      "learning_rate": 0.00017170263860727769,
      "loss": 2.266,
      "step": 1605
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3081947863101959,
      "learning_rate": 0.0001711425646990838,
      "loss": 2.3069,
      "step": 1606
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.31524211168289185,
      "learning_rate": 0.00017058321705965202,
      "loss": 2.2165,
      "step": 1607
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3061792254447937,
      "learning_rate": 0.0001700245969242798,
      "loss": 2.3967,
      "step": 1608
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.31289422512054443,
      "learning_rate": 0.00016946670552665804,
      "loss": 2.292,
      "step": 1609
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.28923431038856506,
      "learning_rate": 0.00016890954409886795,
      "loss": 2.2222,
      "step": 1610
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3235987424850464,
      "learning_rate": 0.00016835311387137836,
      "loss": 2.2945,
      "step": 1611
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.31480979919433594,
      "learning_rate": 0.0001677974160730441,
      "loss": 2.328,
      "step": 1612
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2942544221878052,
      "learning_rate": 0.00016724245193110176,
      "loss": 2.1252,
      "step": 1613
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.31621354818344116,
      "learning_rate": 0.00016668822267116784,
      "loss": 2.2514,
      "step": 1614
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.30238664150238037,
      "learning_rate": 0.00016613472951723597,
      "loss": 2.3675,
      "step": 1615
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3149281442165375,
      "learning_rate": 0.00016558197369167434,
      "loss": 2.3143,
      "step": 1616
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2920989394187927,
      "learning_rate": 0.00016502995641522216,
      "loss": 2.3918,
      "step": 1617
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.30849364399909973,
      "learning_rate": 0.00016447867890698843,
      "loss": 2.367,
      "step": 1618
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3058650493621826,
      "learning_rate": 0.00016392814238444753,
      "loss": 2.2115,
      "step": 1619
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2996141314506531,
      "learning_rate": 0.00016337834806343782,
      "loss": 2.2636,
      "step": 1620
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3069846034049988,
      "learning_rate": 0.0001628292971581588,
      "loss": 2.2772,
      "step": 1621
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2955697774887085,
      "learning_rate": 0.00016228099088116772,
      "loss": 2.3652,
      "step": 1622
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.30348315834999084,
      "learning_rate": 0.00016173343044337734,
      "loss": 2.3135,
      "step": 1623
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.31029775738716125,
      "learning_rate": 0.00016118661705405356,
      "loss": 2.2452,
      "step": 1624
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.2991427481174469,
      "learning_rate": 0.00016064055192081255,
      "loss": 2.3124,
      "step": 1625
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3192315995693207,
      "learning_rate": 0.00016009523624961757,
      "loss": 2.29,
      "step": 1626
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3094353973865509,
      "learning_rate": 0.00015955067124477678,
      "loss": 2.3504,
      "step": 1627
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.31969910860061646,
      "learning_rate": 0.000159006858108941,
      "loss": 2.2074,
      "step": 1628
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.30575475096702576,
      "learning_rate": 0.00015846379804310002,
      "loss": 2.2521,
      "step": 1629
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3111288249492645,
      "learning_rate": 0.00015792149224658054,
      "loss": 2.2413,
      "step": 1630
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3132326602935791,
      "learning_rate": 0.00015737994191704385,
      "loss": 2.216,
      "step": 1631
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3039112985134125,
      "learning_rate": 0.0001568391482504829,
      "loss": 2.2281,
      "step": 1632
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.30036985874176025,
      "learning_rate": 0.00015629911244121903,
      "loss": 2.2301,
      "step": 1633
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.330955445766449,
      "learning_rate": 0.0001557598356819,
      "loss": 2.2151,
      "step": 1634
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3208368718624115,
      "learning_rate": 0.00015522131916349786,
      "loss": 2.2759,
      "step": 1635
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.32030218839645386,
      "learning_rate": 0.00015468356407530493,
      "loss": 2.3029,
      "step": 1636
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.28945404291152954,
      "learning_rate": 0.00015414657160493217,
      "loss": 2.2516,
      "step": 1637
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.2892863154411316,
      "learning_rate": 0.00015361034293830673,
      "loss": 2.256,
      "step": 1638
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.30205604434013367,
      "learning_rate": 0.00015307487925966844,
      "loss": 2.2835,
      "step": 1639
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.297829270362854,
      "learning_rate": 0.00015254018175156776,
      "loss": 2.3387,
      "step": 1640
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.309097021818161,
      "learning_rate": 0.0001520062515948632,
      "loss": 2.2465,
      "step": 1641
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.30266445875167847,
      "learning_rate": 0.0001514730899687189,
      "loss": 2.1741,
      "step": 1642
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.31495994329452515,
      "learning_rate": 0.00015094069805060122,
      "loss": 2.3677,
      "step": 1643
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.314983069896698,
      "learning_rate": 0.00015040907701627666,
      "loss": 2.3238,
      "step": 1644
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.318447470664978,
      "learning_rate": 0.00014987822803980976,
      "loss": 2.3718,
      "step": 1645
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3050040006637573,
      "learning_rate": 0.00014934815229355965,
      "loss": 2.2392,
      "step": 1646
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3068206310272217,
      "learning_rate": 0.00014881885094817748,
      "loss": 2.2597,
      "step": 1647
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.30840086936950684,
      "learning_rate": 0.00014829032517260488,
      "loss": 2.3351,
      "step": 1648
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2977353632450104,
      "learning_rate": 0.0001477625761340704,
      "loss": 2.3315,
      "step": 1649
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.32374706864356995,
      "learning_rate": 0.0001472356049980868,
      "loss": 2.3155,
      "step": 1650
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.4956781268119812,
      "learning_rate": 0.00014670941292844954,
      "loss": 2.2565,
      "step": 1651
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.32776668667793274,
      "learning_rate": 0.00014618400108723295,
      "loss": 2.2114,
      "step": 1652
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3128916323184967,
      "learning_rate": 0.00014565937063478862,
      "loss": 2.2831,
      "step": 1653
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3220754563808441,
      "learning_rate": 0.00014513552272974207,
      "loss": 2.2655,
      "step": 1654
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3159184157848358,
      "learning_rate": 0.0001446124585289913,
      "loss": 2.2637,
      "step": 1655
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.30471664667129517,
      "learning_rate": 0.00014409017918770266,
      "loss": 2.2258,
      "step": 1656
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3254045248031616,
      "learning_rate": 0.00014356868585930994,
      "loss": 2.3672,
      "step": 1657
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3560071289539337,
      "learning_rate": 0.00014304797969551077,
      "loss": 2.2521,
      "step": 1658
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3053613603115082,
      "learning_rate": 0.00014252806184626417,
      "loss": 2.2274,
      "step": 1659
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3017665147781372,
      "learning_rate": 0.00014200893345978817,
      "loss": 2.3159,
      "step": 1660
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3014982044696808,
      "learning_rate": 0.00014149059568255778,
      "loss": 2.1671,
      "step": 1661
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.29759514331817627,
      "learning_rate": 0.00014097304965930157,
      "loss": 2.2308,
      "step": 1662
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.30329740047454834,
      "learning_rate": 0.00014045629653299953,
      "loss": 2.2307,
      "step": 1663
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3063962161540985,
      "learning_rate": 0.00013994033744488076,
      "loss": 2.2209,
      "step": 1664
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.29997608065605164,
      "learning_rate": 0.00013942517353442092,
      "loss": 2.2904,
      "step": 1665
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.33428359031677246,
      "learning_rate": 0.0001389108059393391,
      "loss": 2.4265,
      "step": 1666
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3104774057865143,
      "learning_rate": 0.00013839723579559581,
      "loss": 2.2477,
      "step": 1667
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3157652020454407,
      "learning_rate": 0.00013788446423739103,
      "loss": 2.3401,
      "step": 1668
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.32575851678848267,
      "learning_rate": 0.00013737249239716042,
      "loss": 2.4254,
      "step": 1669
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.31934672594070435,
      "learning_rate": 0.00013686132140557355,
      "loss": 2.2654,
      "step": 1670
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.4972629249095917,
      "learning_rate": 0.00013635095239153188,
      "loss": 2.1756,
      "step": 1671
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.31424471735954285,
      "learning_rate": 0.00013584138648216527,
      "loss": 2.2547,
      "step": 1672
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3186182379722595,
      "learning_rate": 0.0001353326248028298,
      "loss": 2.2712,
      "step": 1673
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.31444185972213745,
      "learning_rate": 0.00013482466847710594,
      "loss": 2.2567,
      "step": 1674
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.31842291355133057,
      "learning_rate": 0.00013431751862679554,
      "loss": 2.287,
      "step": 1675
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3192235231399536,
      "learning_rate": 0.00013381117637191887,
      "loss": 2.3457,
      "step": 1676
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3066616654396057,
      "learning_rate": 0.00013330564283071293,
      "loss": 2.2806,
      "step": 1677
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3336458206176758,
      "learning_rate": 0.000132800919119629,
      "loss": 2.2624,
      "step": 1678
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3116101920604706,
      "learning_rate": 0.00013229700635332948,
      "loss": 2.3224,
      "step": 1679
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.40307390689849854,
      "learning_rate": 0.00013179390564468585,
      "loss": 2.3721,
      "step": 1680
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.5398821830749512,
      "learning_rate": 0.00013129161810477641,
      "loss": 2.326,
      "step": 1681
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3291225731372833,
      "learning_rate": 0.0001307901448428837,
      "loss": 2.3333,
      "step": 1682
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.32499924302101135,
      "learning_rate": 0.0001302894869664916,
      "loss": 2.3498,
      "step": 1683
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3028357923030853,
      "learning_rate": 0.00012978964558128336,
      "loss": 2.2387,
      "step": 1684
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.324585884809494,
      "learning_rate": 0.00012929062179113925,
      "loss": 2.3739,
      "step": 1685
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.320061057806015,
      "learning_rate": 0.00012879241669813368,
      "loss": 2.3456,
      "step": 1686
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.31448403000831604,
      "learning_rate": 0.00012829503140253295,
      "loss": 2.1954,
      "step": 1687
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.30049365758895874,
      "learning_rate": 0.0001277984670027933,
      "loss": 2.2784,
      "step": 1688
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2956079840660095,
      "learning_rate": 0.00012730272459555737,
      "loss": 2.3147,
      "step": 1689
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2984139323234558,
      "learning_rate": 0.00012680780527565312,
      "loss": 2.265,
      "step": 1690
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.307872474193573,
      "learning_rate": 0.0001263137101360905,
      "loss": 2.2999,
      "step": 1691
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.3075066804885864,
      "learning_rate": 0.00012582044026805922,
      "loss": 2.3218,
      "step": 1692
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2980908751487732,
      "learning_rate": 0.00012532799676092627,
      "loss": 2.2449,
      "step": 1693
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.29459989070892334,
      "learning_rate": 0.00012483638070223414,
      "loss": 2.2514,
      "step": 1694
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.31053823232650757,
      "learning_rate": 0.00012434559317769752,
      "loss": 2.2749,
      "step": 1695
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.3120097815990448,
      "learning_rate": 0.0001238556352712012,
      "loss": 2.2369,
      "step": 1696
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.29412853717803955,
      "learning_rate": 0.00012336650806479827,
      "loss": 2.2753,
      "step": 1697
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.31336739659309387,
      "learning_rate": 0.00012287821263870708,
      "loss": 2.3192,
      "step": 1698
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.32532012462615967,
      "learning_rate": 0.00012239075007130885,
      "loss": 2.1936,
      "step": 1699
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.3187802731990814,
      "learning_rate": 0.00012190412143914536,
      "loss": 2.3202,
      "step": 1700
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.30582427978515625,
      "learning_rate": 0.0001214183278169172,
      "loss": 2.3195,
      "step": 1701
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.31001314520835876,
      "learning_rate": 0.00012093337027748042,
      "loss": 2.3567,
      "step": 1702
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.3112240135669708,
      "learning_rate": 0.00012044924989184459,
      "loss": 2.4259,
      "step": 1703
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.3152489960193634,
      "learning_rate": 0.0001199659677291709,
      "loss": 2.3064,
      "step": 1704
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.28679564595222473,
      "learning_rate": 0.00011948352485676895,
      "loss": 2.2406,
      "step": 1705
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.3040367066860199,
      "learning_rate": 0.00011900192234009477,
      "loss": 2.2891,
      "step": 1706
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.30209019780158997,
      "learning_rate": 0.00011852116124274875,
      "loss": 2.223,
      "step": 1707
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.30860963463783264,
      "learning_rate": 0.00011804124262647314,
      "loss": 2.1954,
      "step": 1708
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.3049428164958954,
      "learning_rate": 0.00011756216755114929,
      "loss": 2.1844,
      "step": 1709
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.29445958137512207,
      "learning_rate": 0.00011708393707479548,
      "loss": 2.3142,
      "step": 1710
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.3189711272716522,
      "learning_rate": 0.00011660655225356531,
      "loss": 2.3643,
      "step": 1711
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3116082549095154,
      "learning_rate": 0.0001161300141417444,
      "loss": 2.2833,
      "step": 1712
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.306284099817276,
      "learning_rate": 0.00011565432379174823,
      "loss": 2.3542,
      "step": 1713
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.30264589190483093,
      "learning_rate": 0.00011517948225412056,
      "loss": 2.3269,
      "step": 1714
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.2948421239852905,
      "learning_rate": 0.00011470549057753032,
      "loss": 2.3612,
      "step": 1715
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3039140999317169,
      "learning_rate": 0.00011423234980876957,
      "loss": 2.3087,
      "step": 1716
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.30046430230140686,
      "learning_rate": 0.00011376006099275099,
      "loss": 2.2761,
      "step": 1717
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3007793724536896,
      "learning_rate": 0.00011328862517250609,
      "loss": 2.177,
      "step": 1718
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.29586222767829895,
      "learning_rate": 0.00011281804338918239,
      "loss": 2.2784,
      "step": 1719
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.32645776867866516,
      "learning_rate": 0.00011234831668204115,
      "loss": 2.3544,
      "step": 1720
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.31897640228271484,
      "learning_rate": 0.00011187944608845569,
      "loss": 2.1536,
      "step": 1721
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.29786548018455505,
      "learning_rate": 0.00011141143264390801,
      "loss": 2.341,
      "step": 1722
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3076114356517792,
      "learning_rate": 0.0001109442773819877,
      "loss": 2.249,
      "step": 1723
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.30216822028160095,
      "learning_rate": 0.0001104779813343889,
      "loss": 2.2136,
      "step": 1724
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.31898537278175354,
      "learning_rate": 0.00011001254553090812,
      "loss": 2.3083,
      "step": 1725
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.28576478362083435,
      "learning_rate": 0.00010954797099944186,
      "loss": 2.3197,
      "step": 1726
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3118203580379486,
      "learning_rate": 0.0001090842587659851,
      "loss": 2.3158,
      "step": 1727
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.30730581283569336,
      "learning_rate": 0.00010862140985462804,
      "loss": 2.3377,
      "step": 1728
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3096812665462494,
      "learning_rate": 0.00010815942528755418,
      "loss": 2.3013,
      "step": 1729
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.30246514081954956,
      "learning_rate": 0.00010769830608503844,
      "loss": 2.2567,
      "step": 1730
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.2987140119075775,
      "learning_rate": 0.00010723805326544473,
      "loss": 2.2679,
      "step": 1731
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3135296106338501,
      "learning_rate": 0.00010677866784522316,
      "loss": 2.292,
      "step": 1732
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3158148527145386,
      "learning_rate": 0.00010632015083890839,
      "loss": 2.2194,
      "step": 1733
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2995392680168152,
      "learning_rate": 0.00010586250325911745,
      "loss": 2.3428,
      "step": 1734
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3041720688343048,
      "learning_rate": 0.00010540572611654697,
      "loss": 2.3431,
      "step": 1735
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.31350207328796387,
      "learning_rate": 0.00010494982041997126,
      "loss": 2.2884,
      "step": 1736
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.30291205644607544,
      "learning_rate": 0.0001044947871762405,
      "loss": 2.2335,
      "step": 1737
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3036324381828308,
      "learning_rate": 0.00010404062739027753,
      "loss": 2.2274,
      "step": 1738
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.31174224615097046,
      "learning_rate": 0.00010358734206507641,
      "loss": 2.2029,
      "step": 1739
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.30729690194129944,
      "learning_rate": 0.00010313493220170017,
      "loss": 2.2281,
      "step": 1740
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3073181211948395,
      "learning_rate": 0.00010268339879927836,
      "loss": 2.2503,
      "step": 1741
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.31865209341049194,
      "learning_rate": 0.00010223274285500466,
      "loss": 2.2171,
      "step": 1742
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3078300654888153,
      "learning_rate": 0.00010178296536413495,
      "loss": 2.3351,
      "step": 1743
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.30514851212501526,
      "learning_rate": 0.00010133406731998546,
      "loss": 2.1658,
      "step": 1744
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3127690553665161,
      "learning_rate": 0.00010088604971392979,
      "loss": 2.3009,
      "step": 1745
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2910654842853546,
      "learning_rate": 0.0001004389135353972,
      "loss": 2.1801,
      "step": 1746
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3013940155506134,
      "learning_rate": 9.999265977187049e-05,
      "loss": 2.3775,
      "step": 1747
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.31466981768608093,
      "learning_rate": 9.95472894088838e-05,
      "loss": 2.3468,
      "step": 1748
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3187992572784424,
      "learning_rate": 9.910280343001993e-05,
      "loss": 2.3232,
      "step": 1749
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.31530338525772095,
      "learning_rate": 9.865920281690866e-05,
      "loss": 2.2695,
      "step": 1750
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.31847941875457764,
      "learning_rate": 9.821648854922482e-05,
      "loss": 2.2178,
      "step": 1751
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.30740803480148315,
      "learning_rate": 9.77746616046854e-05,
      "loss": 2.2851,
      "step": 1752
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3024097681045532,
      "learning_rate": 9.733372295904774e-05,
      "loss": 2.3402,
      "step": 1753
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.30137237906455994,
      "learning_rate": 9.68936735861079e-05,
      "loss": 2.2038,
      "step": 1754
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3176831305027008,
      "learning_rate": 9.645451445769737e-05,
      "loss": 2.2696,
      "step": 1755
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.3189190626144409,
      "learning_rate": 9.601624654368196e-05,
      "loss": 2.3196,
      "step": 1756
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.32023051381111145,
      "learning_rate": 9.557887081195938e-05,
      "loss": 2.38,
      "step": 1757
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.321831613779068,
      "learning_rate": 9.514238822845667e-05,
      "loss": 2.325,
      "step": 1758
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.3183562457561493,
      "learning_rate": 9.470679975712837e-05,
      "loss": 2.1841,
      "step": 1759
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.3196760416030884,
      "learning_rate": 9.427210635995481e-05,
      "loss": 2.3103,
      "step": 1760
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.3075222969055176,
      "learning_rate": 9.383830899693923e-05,
      "loss": 2.182,
      "step": 1761
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.29483920335769653,
      "learning_rate": 9.340540862610591e-05,
      "loss": 2.2968,
      "step": 1762
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.31010687351226807,
      "learning_rate": 9.297340620349854e-05,
      "loss": 2.3636,
      "step": 1763
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.314957857131958,
      "learning_rate": 9.25423026831777e-05,
      "loss": 2.2844,
      "step": 1764
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.30944517254829407,
      "learning_rate": 9.211209901721846e-05,
      "loss": 2.2361,
      "step": 1765
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.3236295282840729,
      "learning_rate": 9.168279615570863e-05,
      "loss": 2.2782,
      "step": 1766
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.3270054757595062,
      "learning_rate": 9.125439504674699e-05,
      "loss": 2.3682,
      "step": 1767
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.29893964529037476,
      "learning_rate": 9.082689663644057e-05,
      "loss": 2.3453,
      "step": 1768
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.28749221563339233,
      "learning_rate": 9.040030186890264e-05,
      "loss": 2.2886,
      "step": 1769
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.30487194657325745,
      "learning_rate": 8.997461168625138e-05,
      "loss": 2.2327,
      "step": 1770
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.3268367350101471,
      "learning_rate": 8.954982702860664e-05,
      "loss": 2.2029,
      "step": 1771
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.3120739758014679,
      "learning_rate": 8.912594883408865e-05,
      "loss": 2.265,
      "step": 1772
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.29515549540519714,
      "learning_rate": 8.870297803881589e-05,
      "loss": 2.2425,
      "step": 1773
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.30569973587989807,
      "learning_rate": 8.828091557690287e-05,
      "loss": 2.3478,
      "step": 1774
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.29443275928497314,
      "learning_rate": 8.785976238045801e-05,
      "loss": 2.3659,
      "step": 1775
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.30986326932907104,
      "learning_rate": 8.743951937958144e-05,
      "loss": 2.2795,
      "step": 1776
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3175175189971924,
      "learning_rate": 8.702018750236357e-05,
      "loss": 2.226,
      "step": 1777
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3109266459941864,
      "learning_rate": 8.660176767488237e-05,
      "loss": 2.2089,
      "step": 1778
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3160122334957123,
      "learning_rate": 8.618426082120146e-05,
      "loss": 2.3219,
      "step": 1779
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.311921089887619,
      "learning_rate": 8.576766786336854e-05,
      "loss": 2.3014,
      "step": 1780
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3118340075016022,
      "learning_rate": 8.535198972141294e-05,
      "loss": 2.3723,
      "step": 1781
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3122633993625641,
      "learning_rate": 8.493722731334347e-05,
      "loss": 2.3202,
      "step": 1782
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3135741353034973,
      "learning_rate": 8.452338155514644e-05,
      "loss": 2.0856,
      "step": 1783
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.31397801637649536,
      "learning_rate": 8.411045336078426e-05,
      "loss": 2.313,
      "step": 1784
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.28585362434387207,
      "learning_rate": 8.369844364219264e-05,
      "loss": 2.2984,
      "step": 1785
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.31663987040519714,
      "learning_rate": 8.328735330927873e-05,
      "loss": 2.3201,
      "step": 1786
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3106197118759155,
      "learning_rate": 8.287718326991961e-05,
      "loss": 2.3327,
      "step": 1787
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3107660114765167,
      "learning_rate": 8.246793442995954e-05,
      "loss": 2.2823,
      "step": 1788
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3133731782436371,
      "learning_rate": 8.205960769320875e-05,
      "loss": 2.2555,
      "step": 1789
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3161308765411377,
      "learning_rate": 8.165220396144085e-05,
      "loss": 2.3277,
      "step": 1790
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.29808229207992554,
      "learning_rate": 8.12457241343909e-05,
      "loss": 2.3289,
      "step": 1791
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.30273908376693726,
      "learning_rate": 8.084016910975367e-05,
      "loss": 2.2587,
      "step": 1792
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.29670214653015137,
      "learning_rate": 8.043553978318169e-05,
      "loss": 2.2591,
      "step": 1793
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.29165148735046387,
      "learning_rate": 8.003183704828281e-05,
      "loss": 2.392,
      "step": 1794
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3037622570991516,
      "learning_rate": 7.962906179661872e-05,
      "loss": 2.1704,
      "step": 1795
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.30069372057914734,
      "learning_rate": 7.922721491770296e-05,
      "loss": 2.2457,
      "step": 1796
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3015800714492798,
      "learning_rate": 7.882629729899832e-05,
      "loss": 2.3743,
      "step": 1797
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3229201138019562,
      "learning_rate": 7.842630982591598e-05,
      "loss": 2.4013,
      "step": 1798
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.31271007657051086,
      "learning_rate": 7.802725338181232e-05,
      "loss": 2.3053,
      "step": 1799
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3153951168060303,
      "learning_rate": 7.762912884798812e-05,
      "loss": 2.2369,
      "step": 1800
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3088310658931732,
      "learning_rate": 7.723193710368564e-05,
      "loss": 2.2684,
      "step": 1801
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.31066352128982544,
      "learning_rate": 7.683567902608729e-05,
      "loss": 2.255,
      "step": 1802
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3231235146522522,
      "learning_rate": 7.644035549031364e-05,
      "loss": 2.1696,
      "step": 1803
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.34010186791419983,
      "learning_rate": 7.604596736942115e-05,
      "loss": 2.3647,
      "step": 1804
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.31667059659957886,
      "learning_rate": 7.56525155344004e-05,
      "loss": 2.2675,
      "step": 1805
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.30843329429626465,
      "learning_rate": 7.52600008541745e-05,
      "loss": 2.2991,
      "step": 1806
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3269185721874237,
      "learning_rate": 7.486842419559681e-05,
      "loss": 2.3167,
      "step": 1807
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.29838836193084717,
      "learning_rate": 7.447778642344898e-05,
      "loss": 2.2655,
      "step": 1808
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3166826367378235,
      "learning_rate": 7.408808840043912e-05,
      "loss": 2.2604,
      "step": 1809
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.326072633266449,
      "learning_rate": 7.369933098720021e-05,
      "loss": 2.3274,
      "step": 1810
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.32687848806381226,
      "learning_rate": 7.331151504228767e-05,
      "loss": 2.2385,
      "step": 1811
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3060400187969208,
      "learning_rate": 7.292464142217775e-05,
      "loss": 2.279,
      "step": 1812
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.32335010170936584,
      "learning_rate": 7.25387109812658e-05,
      "loss": 2.3195,
      "step": 1813
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.328336238861084,
      "learning_rate": 7.215372457186415e-05,
      "loss": 2.2631,
      "step": 1814
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3130151629447937,
      "learning_rate": 7.176968304420007e-05,
      "loss": 2.3771,
      "step": 1815
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3256915807723999,
      "learning_rate": 7.138658724641417e-05,
      "loss": 2.2948,
      "step": 1816
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.30590614676475525,
      "learning_rate": 7.10044380245587e-05,
      "loss": 2.2594,
      "step": 1817
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3227904438972473,
      "learning_rate": 7.062323622259515e-05,
      "loss": 2.2968,
      "step": 1818
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3149264454841614,
      "learning_rate": 7.024298268239265e-05,
      "loss": 2.231,
      "step": 1819
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.3041308522224426,
      "learning_rate": 6.986367824372647e-05,
      "loss": 2.3197,
      "step": 1820
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.32574141025543213,
      "learning_rate": 6.948532374427541e-05,
      "loss": 2.2843,
      "step": 1821
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3369005024433136,
      "learning_rate": 6.910792001962063e-05,
      "loss": 2.3001,
      "step": 1822
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.29833143949508667,
      "learning_rate": 6.873146790324358e-05,
      "loss": 2.275,
      "step": 1823
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.30660632252693176,
      "learning_rate": 6.83559682265239e-05,
      "loss": 2.3551,
      "step": 1824
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3223024904727936,
      "learning_rate": 6.798142181873784e-05,
      "loss": 2.2749,
      "step": 1825
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3255276083946228,
      "learning_rate": 6.760782950705662e-05,
      "loss": 2.2509,
      "step": 1826
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.32857921719551086,
      "learning_rate": 6.723519211654422e-05,
      "loss": 2.3919,
      "step": 1827
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.29820355772972107,
      "learning_rate": 6.686351047015554e-05,
      "loss": 2.2799,
      "step": 1828
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3328153192996979,
      "learning_rate": 6.649278538873515e-05,
      "loss": 2.2947,
      "step": 1829
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3038479685783386,
      "learning_rate": 6.612301769101465e-05,
      "loss": 2.2814,
      "step": 1830
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3112033009529114,
      "learning_rate": 6.575420819361177e-05,
      "loss": 2.2337,
      "step": 1831
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3167230486869812,
      "learning_rate": 6.538635771102757e-05,
      "loss": 2.3066,
      "step": 1832
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3118252754211426,
      "learning_rate": 6.501946705564566e-05,
      "loss": 2.3742,
      "step": 1833
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.32644084095954895,
      "learning_rate": 6.465353703772959e-05,
      "loss": 2.3519,
      "step": 1834
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3201438784599304,
      "learning_rate": 6.428856846542136e-05,
      "loss": 2.2895,
      "step": 1835
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3090077340602875,
      "learning_rate": 6.392456214473996e-05,
      "loss": 2.194,
      "step": 1836
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.32453346252441406,
      "learning_rate": 6.3561518879579e-05,
      "loss": 2.357,
      "step": 1837
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3209848403930664,
      "learning_rate": 6.31994394717052e-05,
      "loss": 2.3188,
      "step": 1838
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3040880858898163,
      "learning_rate": 6.283832472075685e-05,
      "loss": 2.3288,
      "step": 1839
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3077160120010376,
      "learning_rate": 6.247817542424178e-05,
      "loss": 2.1129,
      "step": 1840
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.31547027826309204,
      "learning_rate": 6.211899237753559e-05,
      "loss": 2.3358,
      "step": 1841
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.32046011090278625,
      "learning_rate": 6.176077637387984e-05,
      "loss": 2.3435,
      "step": 1842
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.31067755818367004,
      "learning_rate": 6.140352820438066e-05,
      "loss": 2.3196,
      "step": 1843
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.30710986256599426,
      "learning_rate": 6.104724865800665e-05,
      "loss": 2.3669,
      "step": 1844
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.31468796730041504,
      "learning_rate": 6.069193852158711e-05,
      "loss": 2.3343,
      "step": 1845
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.30202698707580566,
      "learning_rate": 6.0337598579810584e-05,
      "loss": 2.3007,
      "step": 1846
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3207736015319824,
      "learning_rate": 5.9984229615223096e-05,
      "loss": 2.2319,
      "step": 1847
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3111916184425354,
      "learning_rate": 5.963183240822606e-05,
      "loss": 2.2683,
      "step": 1848
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3013618588447571,
      "learning_rate": 5.9280407737074825e-05,
      "loss": 2.208,
      "step": 1849
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2905924916267395,
      "learning_rate": 5.8929956377877125e-05,
      "loss": 2.3819,
      "step": 1850
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3189532458782196,
      "learning_rate": 5.8580479104591075e-05,
      "loss": 2.2431,
      "step": 1851
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.30888181924819946,
      "learning_rate": 5.823197668902341e-05,
      "loss": 2.3584,
      "step": 1852
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3128196597099304,
      "learning_rate": 5.78844499008282e-05,
      "loss": 2.2651,
      "step": 1853
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3362501263618469,
      "learning_rate": 5.753789950750454e-05,
      "loss": 2.2531,
      "step": 1854
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3029487729072571,
      "learning_rate": 5.719232627439558e-05,
      "loss": 2.1751,
      "step": 1855
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3130088150501251,
      "learning_rate": 5.6847730964686315e-05,
      "loss": 2.3131,
      "step": 1856
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.30793482065200806,
      "learning_rate": 5.650411433940189e-05,
      "loss": 2.331,
      "step": 1857
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2981932461261749,
      "learning_rate": 5.61614771574061e-05,
      "loss": 2.2526,
      "step": 1858
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.30626237392425537,
      "learning_rate": 5.581982017539988e-05,
      "loss": 2.3173,
      "step": 1859
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.29357489943504333,
      "learning_rate": 5.5479144147919216e-05,
      "loss": 2.1621,
      "step": 1860
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.31725409626960754,
      "learning_rate": 5.51394498273336e-05,
      "loss": 2.2013,
      "step": 1861
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3226546049118042,
      "learning_rate": 5.480073796384494e-05,
      "loss": 2.2353,
      "step": 1862
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3159593641757965,
      "learning_rate": 5.446300930548492e-05,
      "loss": 2.2879,
      "step": 1863
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3364187479019165,
      "learning_rate": 5.412626459811415e-05,
      "loss": 2.2128,
      "step": 1864
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.30622023344039917,
      "learning_rate": 5.3790504585419954e-05,
      "loss": 2.2886,
      "step": 1865
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3081396520137787,
      "learning_rate": 5.345573000891541e-05,
      "loss": 2.2747,
      "step": 1866
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3240078389644623,
      "learning_rate": 5.312194160793693e-05,
      "loss": 2.2954,
      "step": 1867
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.32154950499534607,
      "learning_rate": 5.278914011964303e-05,
      "loss": 2.1526,
      "step": 1868
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2917095124721527,
      "learning_rate": 5.2457326279013006e-05,
      "loss": 2.2709,
      "step": 1869
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3071555197238922,
      "learning_rate": 5.2126500818844514e-05,
      "loss": 2.3014,
      "step": 1870
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3044038712978363,
      "learning_rate": 5.1796664469752566e-05,
      "loss": 2.1223,
      "step": 1871
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3106249272823334,
      "learning_rate": 5.1467817960167975e-05,
      "loss": 2.3261,
      "step": 1872
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.30704057216644287,
      "learning_rate": 5.113996201633536e-05,
      "loss": 2.2351,
      "step": 1873
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.32343819737434387,
      "learning_rate": 5.0813097362311765e-05,
      "loss": 2.3095,
      "step": 1874
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3224557638168335,
      "learning_rate": 5.048722471996475e-05,
      "loss": 2.2822,
      "step": 1875
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3182934522628784,
      "learning_rate": 5.016234480897158e-05,
      "loss": 2.2522,
      "step": 1876
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.31388047337532043,
      "learning_rate": 4.9838458346816664e-05,
      "loss": 2.2138,
      "step": 1877
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2903369069099426,
      "learning_rate": 4.9515566048790485e-05,
      "loss": 2.2579,
      "step": 1878
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.27625587582588196,
      "learning_rate": 4.9193668627988074e-05,
      "loss": 2.1911,
      "step": 1879
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.29623615741729736,
      "learning_rate": 4.887276679530744e-05,
      "loss": 2.2915,
      "step": 1880
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3156781494617462,
      "learning_rate": 4.855286125944752e-05,
      "loss": 2.2862,
      "step": 1881
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.30204516649246216,
      "learning_rate": 4.8233952726907224e-05,
      "loss": 2.1318,
      "step": 1882
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3270680010318756,
      "learning_rate": 4.7916041901983565e-05,
      "loss": 2.3249,
      "step": 1883
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3078593313694,
      "learning_rate": 4.7599129486770145e-05,
      "loss": 2.3122,
      "step": 1884
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3081013560295105,
      "learning_rate": 4.728321618115555e-05,
      "loss": 2.4129,
      "step": 1885
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3100288510322571,
      "learning_rate": 4.696830268282204e-05,
      "loss": 2.3268,
      "step": 1886
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3052492141723633,
      "learning_rate": 4.665438968724361e-05,
      "loss": 2.2701,
      "step": 1887
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.313673198223114,
      "learning_rate": 4.634147788768489e-05,
      "loss": 2.2295,
      "step": 1888
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3047325015068054,
      "learning_rate": 4.6029567975199414e-05,
      "loss": 2.3073,
      "step": 1889
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.29859262704849243,
      "learning_rate": 4.571866063862795e-05,
      "loss": 2.1853,
      "step": 1890
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3154909610748291,
      "learning_rate": 4.540875656459703e-05,
      "loss": 2.3127,
      "step": 1891
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3115561008453369,
      "learning_rate": 4.509985643751785e-05,
      "loss": 2.3035,
      "step": 1892
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3077855408191681,
      "learning_rate": 4.479196093958421e-05,
      "loss": 2.411,
      "step": 1893
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.31866660714149475,
      "learning_rate": 4.4485070750771187e-05,
      "loss": 2.2589,
      "step": 1894
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3143591582775116,
      "learning_rate": 4.417918654883363e-05,
      "loss": 2.2447,
      "step": 1895
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3294619917869568,
      "learning_rate": 4.3874309009305e-05,
      "loss": 2.2003,
      "step": 1896
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.30229100584983826,
      "learning_rate": 4.357043880549538e-05,
      "loss": 2.2811,
      "step": 1897
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3143436014652252,
      "learning_rate": 4.326757660849012e-05,
      "loss": 2.3928,
      "step": 1898
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.31223589181900024,
      "learning_rate": 4.2965723087148635e-05,
      "loss": 2.3239,
      "step": 1899
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2904578447341919,
      "learning_rate": 4.266487890810256e-05,
      "loss": 2.2325,
      "step": 1900
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.31978219747543335,
      "learning_rate": 4.2365044735754365e-05,
      "loss": 2.2156,
      "step": 1901
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.30608394742012024,
      "learning_rate": 4.2066221232276266e-05,
      "loss": 2.4173,
      "step": 1902
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.328331857919693,
      "learning_rate": 4.176840905760815e-05,
      "loss": 2.3475,
      "step": 1903
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.31296029686927795,
      "learning_rate": 4.1471608869456443e-05,
      "loss": 2.2218,
      "step": 1904
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2957618236541748,
      "learning_rate": 4.117582132329284e-05,
      "loss": 2.3453,
      "step": 1905
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3160611093044281,
      "learning_rate": 4.088104707235263e-05,
      "loss": 2.3973,
      "step": 1906
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.30110788345336914,
      "learning_rate": 4.058728676763313e-05,
      "loss": 2.2793,
      "step": 1907
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.31933051347732544,
      "learning_rate": 4.0294541057892375e-05,
      "loss": 2.3549,
      "step": 1908
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.31138473749160767,
      "learning_rate": 4.000281058964794e-05,
      "loss": 2.3334,
      "step": 1909
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.32414019107818604,
      "learning_rate": 3.971209600717507e-05,
      "loss": 2.2109,
      "step": 1910
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3212805390357971,
      "learning_rate": 3.9422397952505465e-05,
      "loss": 2.1862,
      "step": 1911
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.32561731338500977,
      "learning_rate": 3.913371706542596e-05,
      "loss": 2.3211,
      "step": 1912
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3139048218727112,
      "learning_rate": 3.884605398347707e-05,
      "loss": 2.4107,
      "step": 1913
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3155899941921234,
      "learning_rate": 3.8559409341951456e-05,
      "loss": 2.3338,
      "step": 1914
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.30658698081970215,
      "learning_rate": 3.8273783773892404e-05,
      "loss": 2.2731,
      "step": 1915
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.29981905221939087,
      "learning_rate": 3.798917791009293e-05,
      "loss": 2.3254,
      "step": 1916
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2965659499168396,
      "learning_rate": 3.770559237909393e-05,
      "loss": 2.3986,
      "step": 1917
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3218715190887451,
      "learning_rate": 3.742302780718288e-05,
      "loss": 2.2417,
      "step": 1918
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.32250651717185974,
      "learning_rate": 3.7141484818392635e-05,
      "loss": 2.3015,
      "step": 1919
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.31829312443733215,
      "learning_rate": 3.686096403449973e-05,
      "loss": 2.2103,
      "step": 1920
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3105817139148712,
      "learning_rate": 3.658146607502344e-05,
      "loss": 2.2713,
      "step": 1921
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2989356517791748,
      "learning_rate": 3.630299155722411e-05,
      "loss": 2.2649,
      "step": 1922
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.30706679821014404,
      "learning_rate": 3.6025541096101676e-05,
      "loss": 2.3033,
      "step": 1923
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.30304622650146484,
      "learning_rate": 3.574911530439473e-05,
      "loss": 2.2707,
      "step": 1924
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.32507747411727905,
      "learning_rate": 3.5473714792578606e-05,
      "loss": 2.3969,
      "step": 1925
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.31904318928718567,
      "learning_rate": 3.519934016886478e-05,
      "loss": 2.2272,
      "step": 1926
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2976289987564087,
      "learning_rate": 3.4925992039198776e-05,
      "loss": 2.2031,
      "step": 1927
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3141403794288635,
      "learning_rate": 3.465367100725908e-05,
      "loss": 2.2316,
      "step": 1928
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3142629861831665,
      "learning_rate": 3.438237767445618e-05,
      "loss": 2.295,
      "step": 1929
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3073941171169281,
      "learning_rate": 3.4112112639930804e-05,
      "loss": 2.4167,
      "step": 1930
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.31561627984046936,
      "learning_rate": 3.3842876500552564e-05,
      "loss": 2.2934,
      "step": 1931
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3001551032066345,
      "learning_rate": 3.357466985091906e-05,
      "loss": 2.2634,
      "step": 1932
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3222639560699463,
      "learning_rate": 3.330749328335414e-05,
      "loss": 2.3119,
      "step": 1933
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.32129615545272827,
      "learning_rate": 3.304134738790659e-05,
      "loss": 2.2681,
      "step": 1934
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.30402323603630066,
      "learning_rate": 3.277623275234953e-05,
      "loss": 2.1947,
      "step": 1935
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3071843385696411,
      "learning_rate": 3.2512149962177994e-05,
      "loss": 2.2311,
      "step": 1936
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3065943419933319,
      "learning_rate": 3.224909960060851e-05,
      "loss": 2.3549,
      "step": 1937
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.31472674012184143,
      "learning_rate": 3.198708224857755e-05,
      "loss": 2.3154,
      "step": 1938
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3239283561706543,
      "learning_rate": 3.172609848474023e-05,
      "loss": 2.3016,
      "step": 1939
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.31715553998947144,
      "learning_rate": 3.1466148885468895e-05,
      "loss": 2.2209,
      "step": 1940
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3177521824836731,
      "learning_rate": 3.120723402485198e-05,
      "loss": 2.1867,
      "step": 1941
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.31020838022232056,
      "learning_rate": 3.094935447469294e-05,
      "loss": 2.2772,
      "step": 1942
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.31194326281547546,
      "learning_rate": 3.069251080450863e-05,
      "loss": 2.3014,
      "step": 1943
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3095916211605072,
      "learning_rate": 3.0436703581528113e-05,
      "loss": 2.2539,
      "step": 1944
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3084261119365692,
      "learning_rate": 3.0181933370691694e-05,
      "loss": 2.2871,
      "step": 1945
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.29992038011550903,
      "learning_rate": 2.9928200734649523e-05,
      "loss": 2.3151,
      "step": 1946
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3200121521949768,
      "learning_rate": 2.9675506233760142e-05,
      "loss": 2.1611,
      "step": 1947
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.32796141505241394,
      "learning_rate": 2.942385042608925e-05,
      "loss": 2.2519,
      "step": 1948
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3038100302219391,
      "learning_rate": 2.9173233867409054e-05,
      "loss": 2.2329,
      "step": 1949
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3196755647659302,
      "learning_rate": 2.892365711119638e-05,
      "loss": 2.2939,
      "step": 1950
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.3020147681236267,
      "learning_rate": 2.8675120708631596e-05,
      "loss": 2.3792,
      "step": 1951
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3249959945678711,
      "learning_rate": 2.8427625208597764e-05,
      "loss": 2.2737,
      "step": 1952
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3142673075199127,
      "learning_rate": 2.8181171157678874e-05,
      "loss": 2.1595,
      "step": 1953
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.2995107173919678,
      "learning_rate": 2.7935759100159053e-05,
      "loss": 2.3278,
      "step": 1954
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3504796624183655,
      "learning_rate": 2.7691389578021365e-05,
      "loss": 2.2468,
      "step": 1955
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.30841362476348877,
      "learning_rate": 2.7448063130946223e-05,
      "loss": 2.2829,
      "step": 1956
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3157893121242523,
      "learning_rate": 2.7205780296310544e-05,
      "loss": 2.2272,
      "step": 1957
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.31976616382598877,
      "learning_rate": 2.6964541609186378e-05,
      "loss": 2.3655,
      "step": 1958
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.30141764879226685,
      "learning_rate": 2.6724347602340104e-05,
      "loss": 2.2935,
      "step": 1959
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.32281455397605896,
      "learning_rate": 2.6485198806230682e-05,
      "loss": 2.2929,
      "step": 1960
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3133043050765991,
      "learning_rate": 2.6247095749008797e-05,
      "loss": 2.2667,
      "step": 1961
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.31990984082221985,
      "learning_rate": 2.6010038956515826e-05,
      "loss": 2.3181,
      "step": 1962
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3218822479248047,
      "learning_rate": 2.5774028952282423e-05,
      "loss": 2.3417,
      "step": 1963
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3108668625354767,
      "learning_rate": 2.5539066257527277e-05,
      "loss": 2.2768,
      "step": 1964
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3099941313266754,
      "learning_rate": 2.530515139115652e-05,
      "loss": 2.3272,
      "step": 1965
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.296286940574646,
      "learning_rate": 2.5072284869761874e-05,
      "loss": 2.2878,
      "step": 1966
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.5682055950164795,
      "learning_rate": 2.4840467207619786e-05,
      "loss": 2.1846,
      "step": 1967
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.2979433238506317,
      "learning_rate": 2.460969891669068e-05,
      "loss": 2.3085,
      "step": 1968
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3089359998703003,
      "learning_rate": 2.4379980506617272e-05,
      "loss": 2.191,
      "step": 1969
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.32330232858657837,
      "learning_rate": 2.4151312484723464e-05,
      "loss": 2.2518,
      "step": 1970
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3005063533782959,
      "learning_rate": 2.3923695356013798e-05,
      "loss": 2.2391,
      "step": 1971
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3007949888706207,
      "learning_rate": 2.3697129623171833e-05,
      "loss": 2.153,
      "step": 1972
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.29458871483802795,
      "learning_rate": 2.3471615786559042e-05,
      "loss": 2.2624,
      "step": 1973
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.31682559847831726,
      "learning_rate": 2.3247154344213818e-05,
      "loss": 2.3575,
      "step": 1974
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.32115495204925537,
      "learning_rate": 2.3023745791850625e-05,
      "loss": 2.379,
      "step": 1975
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.31377112865448,
      "learning_rate": 2.2801390622858354e-05,
      "loss": 2.2862,
      "step": 1976
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.30361708998680115,
      "learning_rate": 2.2580089328299746e-05,
      "loss": 2.3567,
      "step": 1977
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.3072998523712158,
      "learning_rate": 2.235984239690997e-05,
      "loss": 2.2665,
      "step": 1978
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.3266347646713257,
      "learning_rate": 2.2140650315095934e-05,
      "loss": 2.1435,
      "step": 1979
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2969536781311035,
      "learning_rate": 2.192251356693459e-05,
      "loss": 2.2118,
      "step": 1980
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2971556484699249,
      "learning_rate": 2.170543263417246e-05,
      "loss": 2.2243,
      "step": 1981
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.31474214792251587,
      "learning_rate": 2.1489407996224286e-05,
      "loss": 2.2835,
      "step": 1982
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.30230361223220825,
      "learning_rate": 2.127444013017199e-05,
      "loss": 2.3568,
      "step": 1983
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2962833344936371,
      "learning_rate": 2.1060529510763648e-05,
      "loss": 2.2088,
      "step": 1984
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.3108409345149994,
      "learning_rate": 2.084767661041259e-05,
      "loss": 2.2867,
      "step": 1985
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2996003031730652,
      "learning_rate": 2.063588189919596e-05,
      "loss": 2.2837,
      "step": 1986
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.30516764521598816,
      "learning_rate": 2.0425145844854275e-05,
      "loss": 2.2732,
      "step": 1987
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.31873419880867004,
      "learning_rate": 2.0215468912789693e-05,
      "loss": 2.3362,
      "step": 1988
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.3314947187900543,
      "learning_rate": 2.0006851566065575e-05,
      "loss": 2.3423,
      "step": 1989
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.3135412037372589,
      "learning_rate": 1.9799294265405166e-05,
      "loss": 2.2349,
      "step": 1990
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.32394906878471375,
      "learning_rate": 1.9592797469190572e-05,
      "loss": 2.3741,
      "step": 1991
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.31724119186401367,
      "learning_rate": 1.938736163346194e-05,
      "loss": 2.2359,
      "step": 1992
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.30501222610473633,
      "learning_rate": 1.9182987211916246e-05,
      "loss": 2.3333,
      "step": 1993
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.3058015704154968,
      "learning_rate": 1.8979674655906332e-05,
      "loss": 2.2909,
      "step": 1994
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.30628421902656555,
      "learning_rate": 1.8777424414440024e-05,
      "loss": 2.3087,
      "step": 1995
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.31532570719718933,
      "learning_rate": 1.8576236934179202e-05,
      "loss": 2.2844,
      "step": 1996
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.32065916061401367,
      "learning_rate": 1.8376112659438393e-05,
      "loss": 2.2007,
      "step": 1997
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3165487051010132,
      "learning_rate": 1.8177052032184282e-05,
      "loss": 2.295,
      "step": 1998
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2927776873111725,
      "learning_rate": 1.7979055492034435e-05,
      "loss": 2.1967,
      "step": 1999
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3282099664211273,
      "learning_rate": 1.7782123476256407e-05,
      "loss": 2.279,
      "step": 2000
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.312049001455307,
      "learning_rate": 1.7586256419766965e-05,
      "loss": 2.2371,
      "step": 2001
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3142372667789459,
      "learning_rate": 1.7391454755130766e-05,
      "loss": 2.15,
      "step": 2002
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.323862761259079,
      "learning_rate": 1.7197718912559557e-05,
      "loss": 2.2646,
      "step": 2003
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3153117001056671,
      "learning_rate": 1.700504931991148e-05,
      "loss": 2.2744,
      "step": 2004
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3231937885284424,
      "learning_rate": 1.681344640268978e-05,
      "loss": 2.2399,
      "step": 2005
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.31075626611709595,
      "learning_rate": 1.6622910584041974e-05,
      "loss": 2.2253,
      "step": 2006
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3045026361942291,
      "learning_rate": 1.6433442284758903e-05,
      "loss": 2.181,
      "step": 2007
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3042166829109192,
      "learning_rate": 1.624504192327392e-05,
      "loss": 2.3301,
      "step": 2008
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.32110315561294556,
      "learning_rate": 1.6057709915661856e-05,
      "loss": 2.1857,
      "step": 2009
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3060605823993683,
      "learning_rate": 1.5871446675638057e-05,
      "loss": 2.3254,
      "step": 2010
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3147054612636566,
      "learning_rate": 1.5686252614557638e-05,
      "loss": 2.2753,
      "step": 2011
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.29823437333106995,
      "learning_rate": 1.5502128141414497e-05,
      "loss": 2.0898,
      "step": 2012
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3130391240119934,
      "learning_rate": 1.5319073662840188e-05,
      "loss": 2.3388,
      "step": 2013
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.29129064083099365,
      "learning_rate": 1.5137089583103391e-05,
      "loss": 2.2086,
      "step": 2014
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3061525225639343,
      "learning_rate": 1.4956176304108893e-05,
      "loss": 2.2493,
      "step": 2015
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3068329393863678,
      "learning_rate": 1.4776334225396481e-05,
      "loss": 2.2334,
      "step": 2016
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3038191795349121,
      "learning_rate": 1.4597563744140397e-05,
      "loss": 2.2915,
      "step": 2017
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3229917883872986,
      "learning_rate": 1.4419865255148269e-05,
      "loss": 2.3476,
      "step": 2018
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3087647259235382,
      "learning_rate": 1.4243239150860122e-05,
      "loss": 2.3238,
      "step": 2019
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3097473978996277,
      "learning_rate": 1.4067685821347932e-05,
      "loss": 2.3512,
      "step": 2020
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3052305579185486,
      "learning_rate": 1.389320565431429e-05,
      "loss": 2.3027,
      "step": 2021
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.30837783217430115,
      "learning_rate": 1.3719799035091851e-05,
      "loss": 2.2638,
      "step": 2022
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.31934818625450134,
      "learning_rate": 1.3547466346642278e-05,
      "loss": 2.1923,
      "step": 2023
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.30937033891677856,
      "learning_rate": 1.3376207969555577e-05,
      "loss": 2.2233,
      "step": 2024
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.30508941411972046,
      "learning_rate": 1.32060242820492e-05,
      "loss": 2.2236,
      "step": 2025
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.29451751708984375,
      "learning_rate": 1.3036915659967118e-05,
      "loss": 2.2372,
      "step": 2026
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3045239746570587,
      "learning_rate": 1.2868882476779087e-05,
      "loss": 2.2315,
      "step": 2027
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.31786391139030457,
      "learning_rate": 1.2701925103579815e-05,
      "loss": 2.2213,
      "step": 2028
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.31648319959640503,
      "learning_rate": 1.2536043909088191e-05,
      "loss": 2.2247,
      "step": 2029
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3088999390602112,
      "learning_rate": 1.2371239259646228e-05,
      "loss": 2.1701,
      "step": 2030
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.31133168935775757,
      "learning_rate": 1.2207511519218672e-05,
      "loss": 2.3007,
      "step": 2031
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.31910577416419983,
      "learning_rate": 1.2044861049391676e-05,
      "loss": 2.2482,
      "step": 2032
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.31194379925727844,
      "learning_rate": 1.1883288209372512e-05,
      "loss": 2.3565,
      "step": 2033
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.32084229588508606,
      "learning_rate": 1.1722793355988471e-05,
      "loss": 2.2424,
      "step": 2034
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3074815571308136,
      "learning_rate": 1.1563376843686135e-05,
      "loss": 2.2747,
      "step": 2035
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3162408471107483,
      "learning_rate": 1.140503902453055e-05,
      "loss": 2.2508,
      "step": 2036
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.31786370277404785,
      "learning_rate": 1.1247780248204665e-05,
      "loss": 2.3344,
      "step": 2037
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.299498975276947,
      "learning_rate": 1.1091600862008333e-05,
      "loss": 2.2075,
      "step": 2038
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2986503839492798,
      "learning_rate": 1.0936501210857652e-05,
      "loss": 2.3033,
      "step": 2039
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.31112468242645264,
      "learning_rate": 1.0782481637284013e-05,
      "loss": 2.2628,
      "step": 2040
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.31085485219955444,
      "learning_rate": 1.0629542481433663e-05,
      "loss": 2.248,
      "step": 2041
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3129730820655823,
      "learning_rate": 1.0477684081066751e-05,
      "loss": 2.2456,
      "step": 2042
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3029288947582245,
      "learning_rate": 1.0326906771556566e-05,
      "loss": 2.221,
      "step": 2043
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.315556138753891,
      "learning_rate": 1.017721088588891e-05,
      "loss": 2.3049,
      "step": 2044
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.30684271454811096,
      "learning_rate": 1.0028596754661334e-05,
      "loss": 2.294,
      "step": 2045
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.31256213784217834,
      "learning_rate": 9.881064706082298e-06,
      "loss": 2.267,
      "step": 2046
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.30640295147895813,
      "learning_rate": 9.734615065970454e-06,
      "loss": 2.2802,
      "step": 2047
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3186233639717102,
      "learning_rate": 9.58924815775425e-06,
      "loss": 2.1897,
      "step": 2048
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3060464560985565,
      "learning_rate": 9.444964302470715e-06,
      "loss": 2.3453,
      "step": 2049
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3117521405220032,
      "learning_rate": 9.301763818765018e-06,
      "loss": 2.1742,
      "step": 2050
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.32808658480644226,
      "learning_rate": 9.15964702288996e-06,
      "loss": 2.2098,
      "step": 2051
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.30463486909866333,
      "learning_rate": 9.018614228704925e-06,
      "loss": 2.3548,
      "step": 2052
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3166652023792267,
      "learning_rate": 8.878665747675152e-06,
      "loss": 2.2443,
      "step": 2053
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3349606692790985,
      "learning_rate": 8.739801888871469e-06,
      "loss": 2.313,
      "step": 2054
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.30486544966697693,
      "learning_rate": 8.602022958969336e-06,
      "loss": 2.2851,
      "step": 2055
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3108053207397461,
      "learning_rate": 8.465329262248078e-06,
      "loss": 2.2419,
      "step": 2056
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3104029595851898,
      "learning_rate": 8.32972110059027e-06,
      "loss": 2.3244,
      "step": 2057
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3167918026447296,
      "learning_rate": 8.195198773481406e-06,
      "loss": 2.2524,
      "step": 2058
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2949618995189667,
      "learning_rate": 8.061762578008613e-06,
      "loss": 2.321,
      "step": 2059
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2911970019340515,
      "learning_rate": 7.929412808860559e-06,
      "loss": 2.2467,
      "step": 2060
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.30440205335617065,
      "learning_rate": 7.79814975832649e-06,
      "loss": 2.3063,
      "step": 2061
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.31271064281463623,
      "learning_rate": 7.667973716295851e-06,
      "loss": 2.1878,
      "step": 2062
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3257019519805908,
      "learning_rate": 7.5388849702571205e-06,
      "loss": 2.3123,
      "step": 2063
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3012159466743469,
      "learning_rate": 7.4108838052979185e-06,
      "loss": 2.1839,
      "step": 2064
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3036056160926819,
      "learning_rate": 7.283970504103732e-06,
      "loss": 2.2287,
      "step": 2065
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3124556839466095,
      "learning_rate": 7.1581453469575785e-06,
      "loss": 2.2864,
      "step": 2066
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.30504968762397766,
      "learning_rate": 7.033408611739456e-06,
      "loss": 2.2124,
      "step": 2067
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.30974751710891724,
      "learning_rate": 6.909760573925561e-06,
      "loss": 2.2782,
      "step": 2068
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.297475665807724,
      "learning_rate": 6.787201506587626e-06,
      "loss": 2.1861,
      "step": 2069
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.31659621000289917,
      "learning_rate": 6.66573168039264e-06,
      "loss": 2.1981,
      "step": 2070
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3283754587173462,
      "learning_rate": 6.545351363601959e-06,
      "loss": 2.4562,
      "step": 2071
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.2911643087863922,
      "learning_rate": 6.426060822070812e-06,
      "loss": 2.315,
      "step": 2072
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3179028630256653,
      "learning_rate": 6.3078603192475716e-06,
      "loss": 2.3232,
      "step": 2073
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.31549063324928284,
      "learning_rate": 6.1907501161735934e-06,
      "loss": 2.3077,
      "step": 2074
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.29714900255203247,
      "learning_rate": 6.074730471482049e-06,
      "loss": 2.2847,
      "step": 2075
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.31568002700805664,
      "learning_rate": 5.959801641397755e-06,
      "loss": 2.2835,
      "step": 2076
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3075624406337738,
      "learning_rate": 5.845963879736627e-06,
      "loss": 2.2711,
      "step": 2077
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.2999527156352997,
      "learning_rate": 5.733217437904892e-06,
      "loss": 2.3429,
      "step": 2078
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3292078375816345,
      "learning_rate": 5.621562564898597e-06,
      "loss": 2.2065,
      "step": 2079
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3067309558391571,
      "learning_rate": 5.51099950730316e-06,
      "loss": 2.2341,
      "step": 2080
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3264733850955963,
      "learning_rate": 5.401528509292763e-06,
      "loss": 2.3627,
      "step": 2081
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.31226035952568054,
      "learning_rate": 5.2931498126298495e-06,
      "loss": 2.2896,
      "step": 2082
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.31271734833717346,
      "learning_rate": 5.1858636566645135e-06,
      "loss": 2.3506,
      "step": 2083
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3016485571861267,
      "learning_rate": 5.0796702783340035e-06,
      "loss": 2.3454,
      "step": 2084
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.30729013681411743,
      "learning_rate": 4.97456991216233e-06,
      "loss": 2.2249,
      "step": 2085
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3110896944999695,
      "learning_rate": 4.870562790259325e-06,
      "loss": 2.3235,
      "step": 2086
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.31038618087768555,
      "learning_rate": 4.7676491423208625e-06,
      "loss": 2.1613,
      "step": 2087
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.29248636960983276,
      "learning_rate": 4.66582919562758e-06,
      "loss": 2.2415,
      "step": 2088
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3030036389827728,
      "learning_rate": 4.5651031750448825e-06,
      "loss": 2.2124,
      "step": 2089
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3227144181728363,
      "learning_rate": 4.465471303022217e-06,
      "loss": 2.3181,
      "step": 2090
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.32111015915870667,
      "learning_rate": 4.366933799592743e-06,
      "loss": 2.3474,
      "step": 2091
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3068302273750305,
      "learning_rate": 4.269490882372551e-06,
      "loss": 2.3318,
      "step": 2092
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.30359387397766113,
      "learning_rate": 4.1731427665606115e-06,
      "loss": 2.2242,
      "step": 2093
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.31151530146598816,
      "learning_rate": 4.077889664937884e-06,
      "loss": 2.2448,
      "step": 2094
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.2981361746788025,
      "learning_rate": 3.983731787867207e-06,
      "loss": 2.2891,
      "step": 2095
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3157022297382355,
      "learning_rate": 3.890669343292464e-06,
      "loss": 2.3709,
      "step": 2096
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.32928773760795593,
      "learning_rate": 3.7987025367384743e-06,
      "loss": 2.2007,
      "step": 2097
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.30840009450912476,
      "learning_rate": 3.707831571310327e-06,
      "loss": 2.3562,
      "step": 2098
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3031737208366394,
      "learning_rate": 3.6180566476929912e-06,
      "loss": 2.1855,
      "step": 2099
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.30876755714416504,
      "learning_rate": 3.529377964150815e-06,
      "loss": 2.2795,
      "step": 2100
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3191022276878357,
      "learning_rate": 3.441795716527307e-06,
      "loss": 2.3928,
      "step": 2101
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3122079372406006,
      "learning_rate": 3.355310098244302e-06,
      "loss": 2.2198,
      "step": 2102
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3165730834007263,
      "learning_rate": 3.269921300301959e-06,
      "loss": 2.3115,
      "step": 2103
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.2954213619232178,
      "learning_rate": 3.1856295112780988e-06,
      "loss": 2.2344,
      "step": 2104
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.3033091723918915,
      "learning_rate": 3.102434917327812e-06,
      "loss": 2.2469,
      "step": 2105
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.3113826811313629,
      "learning_rate": 3.0203377021831292e-06,
      "loss": 2.222,
      "step": 2106
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.3097573220729828,
      "learning_rate": 2.939338047152573e-06,
      "loss": 2.1905,
      "step": 2107
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.31251204013824463,
      "learning_rate": 2.8594361311206073e-06,
      "loss": 2.3017,
      "step": 2108
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.31411024928092957,
      "learning_rate": 2.7806321305475225e-06,
      "loss": 2.269,
      "step": 2109
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.31183066964149475,
      "learning_rate": 2.7029262194688818e-06,
      "loss": 2.2219,
      "step": 2110
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.3080286979675293,
      "learning_rate": 2.626318569495134e-06,
      "loss": 2.233,
      "step": 2111
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.31132325530052185,
      "learning_rate": 2.550809349811334e-06,
      "loss": 2.221,
      "step": 2112
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.30469948053359985,
      "learning_rate": 2.476398727176532e-06,
      "loss": 2.1838,
      "step": 2113
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.30013173818588257,
      "learning_rate": 2.4030868659237204e-06,
      "loss": 2.1492,
      "step": 2114
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.3132345378398895,
      "learning_rate": 2.3308739279593317e-06,
      "loss": 2.2379,
      "step": 2115
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.30713897943496704,
      "learning_rate": 2.2597600727626845e-06,
      "loss": 2.2303,
      "step": 2116
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.3196867108345032,
      "learning_rate": 2.1897454573860387e-06,
      "loss": 2.1777,
      "step": 2117
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.32816389203071594,
      "learning_rate": 2.1208302364538746e-06,
      "loss": 2.2372,
      "step": 2118
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.32698407769203186,
      "learning_rate": 2.0530145621627804e-06,
      "loss": 2.1992,
      "step": 2119
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.3028707504272461,
      "learning_rate": 1.9862985842810653e-06,
      "loss": 2.2786,
      "step": 2120
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.3124214708805084,
      "learning_rate": 1.920682450148259e-06,
      "loss": 2.2228,
      "step": 2121
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.31500932574272156,
      "learning_rate": 1.856166304675111e-06,
      "loss": 2.2282,
      "step": 2122
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.2944769561290741,
      "learning_rate": 1.792750290342926e-06,
      "loss": 2.3091,
      "step": 2123
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.30481958389282227,
      "learning_rate": 1.7304345472035632e-06,
      "loss": 2.2873,
      "step": 2124
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.31034135818481445,
      "learning_rate": 1.6692192128788253e-06,
      "loss": 2.2336,
      "step": 2125
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3183976709842682,
      "learning_rate": 1.6091044225604035e-06,
      "loss": 2.2617,
      "step": 2126
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3192707300186157,
      "learning_rate": 1.5500903090094888e-06,
      "loss": 2.2574,
      "step": 2127
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.320273220539093,
      "learning_rate": 1.492177002556383e-06,
      "loss": 2.2923,
      "step": 2128
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.31744661927223206,
      "learning_rate": 1.4353646311004443e-06,
      "loss": 2.2601,
      "step": 2129
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3035826086997986,
      "learning_rate": 1.3796533201094752e-06,
      "loss": 2.1636,
      "step": 2130
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3088661730289459,
      "learning_rate": 1.3250431926197793e-06,
      "loss": 2.2221,
      "step": 2131
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3103650212287903,
      "learning_rate": 1.2715343692356607e-06,
      "loss": 2.3385,
      "step": 2132
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3055790960788727,
      "learning_rate": 1.2191269681292582e-06,
      "loss": 2.2305,
      "step": 2133
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3132423162460327,
      "learning_rate": 1.1678211050402676e-06,
      "loss": 2.2793,
      "step": 2134
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.32375580072402954,
      "learning_rate": 1.117616893275719e-06,
      "loss": 2.3257,
      "step": 2135
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.31228604912757874,
      "learning_rate": 1.068514443709534e-06,
      "loss": 2.3015,
      "step": 2136
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3016040027141571,
      "learning_rate": 1.0205138647826905e-06,
      "loss": 2.2833,
      "step": 2137
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3170969784259796,
      "learning_rate": 9.73615262502503e-07,
      "loss": 2.2461,
      "step": 2138
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3219013810157776,
      "learning_rate": 9.278187404426763e-07,
      "loss": 2.2937,
      "step": 2139
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.30782783031463623,
      "learning_rate": 8.831243997431404e-07,
      "loss": 2.2551,
      "step": 2140
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3126419484615326,
      "learning_rate": 8.395323391094944e-07,
      "loss": 2.3232,
      "step": 2141
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.32309073209762573,
      "learning_rate": 7.970426548131183e-07,
      "loss": 2.2465,
      "step": 2142
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3140074610710144,
      "learning_rate": 7.556554406908389e-07,
      "loss": 2.2674,
      "step": 2143
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3025936484336853,
      "learning_rate": 7.153707881446536e-07,
      "loss": 2.3295,
      "step": 2144
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3132850229740143,
      "learning_rate": 6.761887861417293e-07,
      "loss": 2.2081,
      "step": 2145
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3029237985610962,
      "learning_rate": 6.381095212139032e-07,
      "loss": 2.3083,
      "step": 2146
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.3048817813396454,
      "learning_rate": 6.011330774577384e-07,
      "loss": 2.3559,
      "step": 2147
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3104776442050934,
      "learning_rate": 5.652595365343016e-07,
      "loss": 2.3969,
      "step": 2148
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3305276930332184,
      "learning_rate": 5.304889776688859e-07,
      "loss": 2.2492,
      "step": 2149
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.29690855741500854,
      "learning_rate": 4.968214776508994e-07,
      "loss": 2.2036,
      "step": 2150
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.31141197681427,
      "learning_rate": 4.6425711083375454e-07,
      "loss": 2.198,
      "step": 2151
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3060808777809143,
      "learning_rate": 4.3279594913447906e-07,
      "loss": 2.295,
      "step": 2152
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.31115588545799255,
      "learning_rate": 4.02438062033883e-07,
      "loss": 2.2816,
      "step": 2153
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3203050196170807,
      "learning_rate": 3.7318351657616987e-07,
      "loss": 2.1839,
      "step": 2154
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.2971652150154114,
      "learning_rate": 3.4503237736882573e-07,
      "loss": 2.2604,
      "step": 2155
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3194325864315033,
      "learning_rate": 3.179847065825081e-07,
      "loss": 2.3096,
      "step": 2156
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3085038363933563,
      "learning_rate": 2.9204056395104594e-07,
      "loss": 2.1471,
      "step": 2157
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.30935269594192505,
      "learning_rate": 2.672000067709956e-07,
      "loss": 2.1664,
      "step": 2158
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3095177710056305,
      "learning_rate": 2.4346308990175204e-07,
      "loss": 2.3326,
      "step": 2159
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.30707207322120667,
      "learning_rate": 2.208298657653818e-07,
      "loss": 2.2979,
      "step": 2160
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.32870161533355713,
      "learning_rate": 1.9930038434645692e-07,
      "loss": 2.3904,
      "step": 2161
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3136035203933716,
      "learning_rate": 1.7887469319205484e-07,
      "loss": 2.2665,
      "step": 2162
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.32743483781814575,
      "learning_rate": 1.5955283741142523e-07,
      "loss": 2.2591,
      "step": 2163
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3139297068119049,
      "learning_rate": 1.4133485967615655e-07,
      "loss": 2.3046,
      "step": 2164
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3242689073085785,
      "learning_rate": 1.2422080021995407e-07,
      "loss": 2.3341,
      "step": 2165
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.31361934542655945,
      "learning_rate": 1.082106968385288e-07,
      "loss": 2.2417,
      "step": 2166
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3216847777366638,
      "learning_rate": 9.330458488959748e-08,
      "loss": 2.3321,
      "step": 2167
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3099242150783539,
      "learning_rate": 7.950249729271608e-08,
      "loss": 2.3419,
      "step": 2168
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.3097037076950073,
      "learning_rate": 6.680446452922429e-08,
      "loss": 2.176,
      "step": 2169
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3216911554336548,
      "learning_rate": 5.521051464230098e-08,
      "loss": 2.1692,
      "step": 2170
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3193168044090271,
      "learning_rate": 4.4720673236631206e-08,
      "loss": 2.2994,
      "step": 2171
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2940231263637543,
      "learning_rate": 3.53349634786837e-08,
      "loss": 2.1813,
      "step": 2172
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2977256178855896,
      "learning_rate": 2.7053406096433365e-08,
      "loss": 2.2082,
      "step": 2173
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.30933254957199097,
      "learning_rate": 1.987601937930572e-08,
      "loss": 2.2444,
      "step": 2174
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.31380873918533325,
      "learning_rate": 1.3802819178398984e-08,
      "loss": 2.3465,
      "step": 2175
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3046720027923584,
      "learning_rate": 8.833818906039959e-09,
      "loss": 2.2448,
      "step": 2176
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3153587579727173,
      "learning_rate": 4.969029536061598e-09,
      "loss": 2.3466,
      "step": 2177
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3061943054199219,
      "learning_rate": 2.2084596038030037e-09,
      "loss": 2.1764,
      "step": 2178
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2996554672718048,
      "learning_rate": 5.521152057763601e-10,
      "loss": 2.2046,
      "step": 2179
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.32201087474823,
      "learning_rate": 0.0,
      "loss": 2.4262,
      "step": 2180
    },
    {
      "epoch": 1.0,
      "step": 2180,
      "total_flos": 5.5653188118144614e+17,
      "train_loss": 2.478320124280562,
      "train_runtime": 41033.5864,
      "train_samples_per_second": 13.602,
      "train_steps_per_second": 0.053
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 2180,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 400,
  "total_flos": 5.5653188118144614e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}