{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.01,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1e-05,
      "grad_norm": 1.085229352812366,
      "learning_rate": 3e-06,
      "loss": 10.849,
      "step": 1
    },
    {
      "epoch": 2e-05,
      "grad_norm": 1.0764689186661929,
      "learning_rate": 6e-06,
      "loss": 10.8489,
      "step": 2
    },
    {
      "epoch": 3e-05,
      "grad_norm": 1.0926036068515363,
      "learning_rate": 9e-06,
      "loss": 10.8486,
      "step": 3
    },
    {
      "epoch": 4e-05,
      "grad_norm": 1.0859011783792423,
      "learning_rate": 1.2e-05,
      "loss": 10.848,
      "step": 4
    },
    {
      "epoch": 5e-05,
      "grad_norm": 1.0906873388641662,
      "learning_rate": 1.5e-05,
      "loss": 10.8453,
      "step": 5
    },
    {
      "epoch": 6e-05,
      "grad_norm": 1.0895888734627917,
      "learning_rate": 1.8e-05,
      "loss": 10.8447,
      "step": 6
    },
    {
      "epoch": 7e-05,
      "grad_norm": 1.0913329404483254,
      "learning_rate": 2.1000000000000002e-05,
      "loss": 10.8355,
      "step": 7
    },
    {
      "epoch": 8e-05,
      "grad_norm": 1.0766237663279077,
      "learning_rate": 2.4e-05,
      "loss": 10.8141,
      "step": 8
    },
    {
      "epoch": 9e-05,
      "grad_norm": 1.0617425338278697,
      "learning_rate": 2.7e-05,
      "loss": 10.8099,
      "step": 9
    },
    {
      "epoch": 0.0001,
      "grad_norm": 1.080271312928173,
      "learning_rate": 3e-05,
      "loss": 10.7983,
      "step": 10
    },
    {
      "epoch": 0.00011,
      "grad_norm": 1.0663119503184246,
      "learning_rate": 3.2999999999999996e-05,
      "loss": 10.7844,
      "step": 11
    },
    {
      "epoch": 0.00012,
      "grad_norm": 1.0694505502575085,
      "learning_rate": 3.6e-05,
      "loss": 10.7739,
      "step": 12
    },
    {
      "epoch": 0.00013,
      "grad_norm": 1.0509206118753271,
      "learning_rate": 3.9e-05,
      "loss": 10.7554,
      "step": 13
    },
    {
      "epoch": 0.00014,
      "grad_norm": 1.0427932624919014,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 10.7452,
      "step": 14
    },
    {
      "epoch": 0.00015,
      "grad_norm": 1.0323600599358198,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 10.736,
      "step": 15
    },
    {
      "epoch": 0.00016,
      "grad_norm": 1.0039727604034705,
      "learning_rate": 4.8e-05,
      "loss": 10.7191,
      "step": 16
    },
    {
      "epoch": 0.00017,
      "grad_norm": 0.9736033767894897,
      "learning_rate": 5.1000000000000006e-05,
      "loss": 10.7049,
      "step": 17
    },
    {
      "epoch": 0.00018,
      "grad_norm": 0.9665971933165531,
      "learning_rate": 5.4e-05,
      "loss": 10.69,
      "step": 18
    },
    {
      "epoch": 0.00019,
      "grad_norm": 0.9428089164290921,
      "learning_rate": 5.7e-05,
      "loss": 10.6767,
      "step": 19
    },
    {
      "epoch": 0.0002,
      "grad_norm": 0.943192779381936,
      "learning_rate": 6e-05,
      "loss": 10.666,
      "step": 20
    },
    {
      "epoch": 0.00021,
      "grad_norm": 0.9246956411226879,
      "learning_rate": 6.3e-05,
      "loss": 10.6525,
      "step": 21
    },
    {
      "epoch": 0.00022,
      "grad_norm": 0.9206776841693647,
      "learning_rate": 6.599999999999999e-05,
      "loss": 10.6378,
      "step": 22
    },
    {
      "epoch": 0.00023,
      "grad_norm": 0.9152142474292139,
      "learning_rate": 6.9e-05,
      "loss": 10.626,
      "step": 23
    },
    {
      "epoch": 0.00024,
      "grad_norm": 0.9101286615991204,
      "learning_rate": 7.2e-05,
      "loss": 10.613,
      "step": 24
    },
    {
      "epoch": 0.00025,
      "grad_norm": 0.9100685942512898,
      "learning_rate": 7.500000000000001e-05,
      "loss": 10.5998,
      "step": 25
    },
    {
      "epoch": 0.00026,
      "grad_norm": 0.9093157110032968,
      "learning_rate": 7.8e-05,
      "loss": 10.5859,
      "step": 26
    },
    {
      "epoch": 0.00027,
      "grad_norm": 0.9036082412984342,
      "learning_rate": 8.1e-05,
      "loss": 10.5742,
      "step": 27
    },
    {
      "epoch": 0.00028,
      "grad_norm": 0.9033464449252919,
      "learning_rate": 8.400000000000001e-05,
      "loss": 10.5612,
      "step": 28
    },
    {
      "epoch": 0.00029,
      "grad_norm": 0.9067163029768021,
      "learning_rate": 8.7e-05,
      "loss": 10.5465,
      "step": 29
    },
    {
      "epoch": 0.0003,
      "grad_norm": 0.9152916501496612,
      "learning_rate": 8.999999999999999e-05,
      "loss": 10.5321,
      "step": 30
    },
    {
      "epoch": 0.00031,
      "grad_norm": 0.9122391999445313,
      "learning_rate": 9.3e-05,
      "loss": 10.5167,
      "step": 31
    },
    {
      "epoch": 0.00032,
      "grad_norm": 0.9090668181794543,
      "learning_rate": 9.6e-05,
      "loss": 10.5034,
      "step": 32
    },
    {
      "epoch": 0.00033,
      "grad_norm": 0.9095629132760054,
      "learning_rate": 9.900000000000001e-05,
      "loss": 10.4865,
      "step": 33
    },
    {
      "epoch": 0.00034,
      "grad_norm": 0.9107692784496799,
      "learning_rate": 0.00010200000000000001,
      "loss": 10.4706,
      "step": 34
    },
    {
      "epoch": 0.00035,
      "grad_norm": 0.9162740070354684,
      "learning_rate": 0.00010500000000000002,
      "loss": 10.4535,
      "step": 35
    },
    {
      "epoch": 0.00036,
      "grad_norm": 0.9114324016484531,
      "learning_rate": 0.000108,
      "loss": 10.4367,
      "step": 36
    },
    {
      "epoch": 0.00037,
      "grad_norm": 0.901331547883087,
      "learning_rate": 0.000111,
      "loss": 10.4196,
      "step": 37
    },
    {
      "epoch": 0.00038,
      "grad_norm": 0.9056352937834914,
      "learning_rate": 0.000114,
      "loss": 10.4003,
      "step": 38
    },
    {
      "epoch": 0.00039,
      "grad_norm": 0.9071824304759276,
      "learning_rate": 0.000117,
      "loss": 10.3815,
      "step": 39
    },
    {
      "epoch": 0.0004,
      "grad_norm": 0.911150856035614,
      "learning_rate": 0.00012,
      "loss": 10.3599,
      "step": 40
    },
    {
      "epoch": 0.00041,
      "grad_norm": 0.9001357892993758,
      "learning_rate": 0.000123,
      "loss": 10.3411,
      "step": 41
    },
    {
      "epoch": 0.00042,
      "grad_norm": 0.9057215347961051,
      "learning_rate": 0.000126,
      "loss": 10.3208,
      "step": 42
    },
    {
      "epoch": 0.00043,
      "grad_norm": 0.9101872556211666,
      "learning_rate": 0.000129,
      "loss": 10.2978,
      "step": 43
    },
    {
      "epoch": 0.00044,
      "grad_norm": 0.9097345366629672,
      "learning_rate": 0.00013199999999999998,
      "loss": 10.2768,
      "step": 44
    },
    {
      "epoch": 0.00045,
      "grad_norm": 0.9111670268341607,
      "learning_rate": 0.000135,
      "loss": 10.2539,
      "step": 45
    },
    {
      "epoch": 0.00046,
      "grad_norm": 0.9030473796809102,
      "learning_rate": 0.000138,
      "loss": 10.2322,
      "step": 46
    },
    {
      "epoch": 0.00047,
      "grad_norm": 0.9086055117133346,
      "learning_rate": 0.000141,
      "loss": 10.2064,
      "step": 47
    },
    {
      "epoch": 0.00048,
      "grad_norm": 0.9060414917101882,
      "learning_rate": 0.000144,
      "loss": 10.1837,
      "step": 48
    },
    {
      "epoch": 0.00049,
      "grad_norm": 0.9059366436676172,
      "learning_rate": 0.000147,
      "loss": 10.1599,
      "step": 49
    },
    {
      "epoch": 0.0005,
      "grad_norm": 0.9106165244124662,
      "learning_rate": 0.00015000000000000001,
      "loss": 10.1343,
      "step": 50
    },
    {
      "epoch": 0.00051,
      "grad_norm": 0.9110992341155927,
      "learning_rate": 0.000153,
      "loss": 10.1078,
      "step": 51
    },
    {
      "epoch": 0.00052,
      "grad_norm": 0.9077294523989683,
      "learning_rate": 0.000156,
      "loss": 10.0815,
      "step": 52
    },
    {
      "epoch": 0.00053,
      "grad_norm": 0.8995624264210066,
      "learning_rate": 0.000159,
      "loss": 10.0581,
      "step": 53
    },
    {
      "epoch": 0.00054,
      "grad_norm": 0.9187536344258231,
      "learning_rate": 0.000162,
      "loss": 10.026,
      "step": 54
    },
    {
      "epoch": 0.00055,
      "grad_norm": 0.9102867236672618,
      "learning_rate": 0.000165,
      "loss": 10.0019,
      "step": 55
    },
    {
      "epoch": 0.00056,
      "grad_norm": 0.9031170239719724,
      "learning_rate": 0.00016800000000000002,
      "loss": 9.9743,
      "step": 56
    },
    {
      "epoch": 0.00057,
      "grad_norm": 0.9090239107255728,
      "learning_rate": 0.000171,
      "loss": 9.9467,
      "step": 57
    },
    {
      "epoch": 0.00058,
      "grad_norm": 0.9070896749665766,
      "learning_rate": 0.000174,
      "loss": 9.9223,
      "step": 58
    },
    {
      "epoch": 0.00059,
      "grad_norm": 0.9116582783399498,
      "learning_rate": 0.000177,
      "loss": 9.8905,
      "step": 59
    },
    {
      "epoch": 0.0006,
      "grad_norm": 0.899071209460366,
      "learning_rate": 0.00017999999999999998,
      "loss": 9.8642,
      "step": 60
    },
    {
      "epoch": 0.00061,
      "grad_norm": 0.8984032767613607,
      "learning_rate": 0.000183,
      "loss": 9.8367,
      "step": 61
    },
    {
      "epoch": 0.00062,
      "grad_norm": 0.9043779152804675,
      "learning_rate": 0.000186,
      "loss": 9.8058,
      "step": 62
    },
    {
      "epoch": 0.00063,
      "grad_norm": 0.8969230453763369,
      "learning_rate": 0.000189,
      "loss": 9.7784,
      "step": 63
    },
    {
      "epoch": 0.00064,
      "grad_norm": 0.8929896633083918,
      "learning_rate": 0.000192,
      "loss": 9.7485,
      "step": 64
    },
    {
      "epoch": 0.00065,
      "grad_norm": 0.9016512171852502,
      "learning_rate": 0.00019500000000000002,
      "loss": 9.7165,
      "step": 65
    },
    {
      "epoch": 0.00066,
      "grad_norm": 0.8946267084059,
      "learning_rate": 0.00019800000000000002,
      "loss": 9.6927,
      "step": 66
    },
    {
      "epoch": 0.00067,
      "grad_norm": 0.8963872265737496,
      "learning_rate": 0.000201,
      "loss": 9.6633,
      "step": 67
    },
    {
      "epoch": 0.00068,
      "grad_norm": 0.9035635569767697,
      "learning_rate": 0.00020400000000000003,
      "loss": 9.6313,
      "step": 68
    },
    {
      "epoch": 0.00069,
      "grad_norm": 0.8884430485081615,
      "learning_rate": 0.00020700000000000002,
      "loss": 9.604,
      "step": 69
    },
    {
      "epoch": 0.0007,
      "grad_norm": 0.901710386427562,
      "learning_rate": 0.00021000000000000004,
      "loss": 9.5675,
      "step": 70
    },
    {
      "epoch": 0.00071,
      "grad_norm": 0.8913815692585527,
      "learning_rate": 0.00021299999999999997,
      "loss": 9.5403,
      "step": 71
    },
    {
      "epoch": 0.00072,
      "grad_norm": 0.8947982003050186,
      "learning_rate": 0.000216,
      "loss": 9.5138,
      "step": 72
    },
    {
      "epoch": 0.00073,
      "grad_norm": 0.8936046055705469,
      "learning_rate": 0.00021899999999999998,
      "loss": 9.4802,
      "step": 73
    },
    {
      "epoch": 0.00074,
      "grad_norm": 0.894403551075387,
      "learning_rate": 0.000222,
      "loss": 9.4539,
      "step": 74
    },
    {
      "epoch": 0.00075,
      "grad_norm": 0.8897139679621429,
      "learning_rate": 0.000225,
      "loss": 9.4187,
      "step": 75
    },
    {
      "epoch": 0.00076,
      "grad_norm": 0.89755782108852,
      "learning_rate": 0.000228,
      "loss": 9.3929,
      "step": 76
    },
    {
      "epoch": 0.00077,
      "grad_norm": 0.8917412779828411,
      "learning_rate": 0.000231,
      "loss": 9.3649,
      "step": 77
    },
    {
      "epoch": 0.00078,
      "grad_norm": 0.886930008239094,
      "learning_rate": 0.000234,
      "loss": 9.3351,
      "step": 78
    },
    {
      "epoch": 0.00079,
      "grad_norm": 0.8893115023720741,
      "learning_rate": 0.00023700000000000001,
      "loss": 9.2992,
      "step": 79
    },
    {
      "epoch": 0.0008,
      "grad_norm": 0.8917772276535281,
      "learning_rate": 0.00024,
      "loss": 9.2632,
      "step": 80
    },
    {
      "epoch": 0.00081,
      "grad_norm": 0.8971231461853929,
      "learning_rate": 0.00024300000000000002,
      "loss": 9.2316,
      "step": 81
    },
    {
      "epoch": 0.00082,
      "grad_norm": 0.9071155901294882,
      "learning_rate": 0.000246,
      "loss": 9.206,
      "step": 82
    },
    {
      "epoch": 0.00083,
      "grad_norm": 0.8963243651316662,
      "learning_rate": 0.00024900000000000004,
      "loss": 9.1708,
      "step": 83
    },
    {
      "epoch": 0.00084,
      "grad_norm": 0.9002519637093493,
      "learning_rate": 0.000252,
      "loss": 9.1395,
      "step": 84
    },
    {
      "epoch": 0.00085,
      "grad_norm": 0.8962870771351267,
      "learning_rate": 0.000255,
      "loss": 9.1159,
      "step": 85
    },
    {
      "epoch": 0.00086,
      "grad_norm": 0.8973951423301171,
      "learning_rate": 0.000258,
      "loss": 9.0799,
      "step": 86
    },
    {
      "epoch": 0.00087,
      "grad_norm": 0.8894399774297843,
      "learning_rate": 0.000261,
      "loss": 9.0571,
      "step": 87
    },
    {
      "epoch": 0.00088,
      "grad_norm": 0.8954974565265822,
      "learning_rate": 0.00026399999999999997,
      "loss": 9.0211,
      "step": 88
    },
    {
      "epoch": 0.00089,
      "grad_norm": 0.8903887799592864,
      "learning_rate": 0.000267,
      "loss": 8.9943,
      "step": 89
    },
    {
      "epoch": 0.0009,
      "grad_norm": 0.8896831425922581,
      "learning_rate": 0.00027,
      "loss": 8.9666,
      "step": 90
    },
    {
      "epoch": 0.00091,
      "grad_norm": 0.8841935030693385,
      "learning_rate": 0.000273,
      "loss": 8.9365,
      "step": 91
    },
    {
      "epoch": 0.00092,
      "grad_norm": 0.8862593694142118,
      "learning_rate": 0.000276,
      "loss": 8.9052,
      "step": 92
    },
    {
      "epoch": 0.00093,
      "grad_norm": 0.8881041222250594,
      "learning_rate": 0.000279,
      "loss": 8.8752,
      "step": 93
    },
    {
      "epoch": 0.00094,
      "grad_norm": 0.8868931156100198,
      "learning_rate": 0.000282,
      "loss": 8.8494,
      "step": 94
    },
    {
      "epoch": 0.00095,
      "grad_norm": 0.8826333164427848,
      "learning_rate": 0.000285,
      "loss": 8.8203,
      "step": 95
    },
    {
      "epoch": 0.00096,
      "grad_norm": 0.8910955494857569,
      "learning_rate": 0.000288,
      "loss": 8.7905,
      "step": 96
    },
    {
      "epoch": 0.00097,
      "grad_norm": 0.8809279978723125,
      "learning_rate": 0.000291,
      "loss": 8.7668,
      "step": 97
    },
    {
      "epoch": 0.00098,
      "grad_norm": 0.8829202734417477,
      "learning_rate": 0.000294,
      "loss": 8.73,
      "step": 98
    },
    {
      "epoch": 0.00099,
      "grad_norm": 0.8858154061322314,
      "learning_rate": 0.000297,
      "loss": 8.7072,
      "step": 99
    },
    {
      "epoch": 0.001,
      "grad_norm": 0.8861735275161908,
      "learning_rate": 0.00030000000000000003,
      "loss": 8.6806,
      "step": 100
    },
    {
      "epoch": 0.00101,
      "grad_norm": 0.8886468466177152,
      "learning_rate": 0.00030300000000000005,
      "loss": 8.6455,
      "step": 101
    },
    {
      "epoch": 0.00102,
      "grad_norm": 0.8755547614914917,
      "learning_rate": 0.000306,
      "loss": 8.625,
      "step": 102
    },
    {
      "epoch": 0.00103,
      "grad_norm": 0.879237266457317,
      "learning_rate": 0.000309,
      "loss": 8.6003,
      "step": 103
    },
    {
      "epoch": 0.00104,
      "grad_norm": 0.8697854479956653,
      "learning_rate": 0.000312,
      "loss": 8.5786,
      "step": 104
    },
    {
      "epoch": 0.00105,
      "grad_norm": 0.8733949043334459,
      "learning_rate": 0.000315,
      "loss": 8.552,
      "step": 105
    },
    {
      "epoch": 0.00106,
      "grad_norm": 0.8706148334292045,
      "learning_rate": 0.000318,
      "loss": 8.5224,
      "step": 106
    },
    {
      "epoch": 0.00107,
      "grad_norm": 0.8653683244965015,
      "learning_rate": 0.000321,
      "loss": 8.4981,
      "step": 107
    },
    {
      "epoch": 0.00108,
      "grad_norm": 0.8737239863866451,
      "learning_rate": 0.000324,
      "loss": 8.4698,
      "step": 108
    },
    {
      "epoch": 0.00109,
      "grad_norm": 0.8706512729216435,
      "learning_rate": 0.000327,
      "loss": 8.4501,
      "step": 109
    },
    {
      "epoch": 0.0011,
      "grad_norm": 0.8842127152624679,
      "learning_rate": 0.00033,
      "loss": 8.4274,
      "step": 110
    },
    {
      "epoch": 0.00111,
      "grad_norm": 0.9007754832304464,
      "learning_rate": 0.000333,
      "loss": 8.3985,
      "step": 111
    },
    {
      "epoch": 0.00112,
      "grad_norm": 0.9159855921530741,
      "learning_rate": 0.00033600000000000004,
      "loss": 8.3784,
      "step": 112
    },
    {
      "epoch": 0.00113,
      "grad_norm": 0.9063278036144603,
      "learning_rate": 0.000339,
      "loss": 8.3391,
      "step": 113
    },
    {
      "epoch": 0.00114,
      "grad_norm": 0.8437820836704115,
      "learning_rate": 0.000342,
      "loss": 8.3286,
      "step": 114
    },
    {
      "epoch": 0.00115,
      "grad_norm": 0.8612821674982505,
      "learning_rate": 0.00034500000000000004,
      "loss": 8.3022,
      "step": 115
    },
    {
      "epoch": 0.00116,
      "grad_norm": 0.8581012057508914,
      "learning_rate": 0.000348,
      "loss": 8.2786,
      "step": 116
    },
    {
      "epoch": 0.00117,
      "grad_norm": 0.834681162463853,
      "learning_rate": 0.000351,
      "loss": 8.2472,
      "step": 117
    },
    {
      "epoch": 0.00118,
      "grad_norm": 0.8526508674143746,
      "learning_rate": 0.000354,
      "loss": 8.2312,
      "step": 118
    },
    {
      "epoch": 0.00119,
      "grad_norm": 0.8471842679679056,
      "learning_rate": 0.000357,
      "loss": 8.2186,
      "step": 119
    },
    {
      "epoch": 0.0012,
      "grad_norm": 0.8238029079166322,
      "learning_rate": 0.00035999999999999997,
      "loss": 8.1917,
      "step": 120
    },
    {
      "epoch": 0.00121,
      "grad_norm": 0.832770075662114,
      "learning_rate": 0.000363,
      "loss": 8.1641,
      "step": 121
    },
    {
      "epoch": 0.00122,
      "grad_norm": 0.8176689288160716,
      "learning_rate": 0.000366,
      "loss": 8.1471,
      "step": 122
    },
    {
      "epoch": 0.00123,
      "grad_norm": 0.8121975389696077,
      "learning_rate": 0.000369,
      "loss": 8.1243,
      "step": 123
    },
    {
      "epoch": 0.00124,
      "grad_norm": 0.8287555396618358,
      "learning_rate": 0.000372,
      "loss": 8.1037,
      "step": 124
    },
    {
      "epoch": 0.00125,
      "grad_norm": 0.8387006564379252,
      "learning_rate": 0.000375,
      "loss": 8.0802,
      "step": 125
    },
    {
      "epoch": 0.00126,
      "grad_norm": 0.7985855201767323,
      "learning_rate": 0.000378,
      "loss": 8.0672,
      "step": 126
    },
    {
      "epoch": 0.00127,
      "grad_norm": 0.7830434817433392,
      "learning_rate": 0.000381,
      "loss": 8.0437,
      "step": 127
    },
    {
      "epoch": 0.00128,
      "grad_norm": 0.8466276209824322,
      "learning_rate": 0.000384,
      "loss": 8.0264,
      "step": 128
    },
    {
      "epoch": 0.00129,
      "grad_norm": 1.1259201196462498,
      "learning_rate": 0.00038700000000000003,
      "loss": 8.0199,
      "step": 129
    },
    {
      "epoch": 0.0013,
      "grad_norm": 1.356061683657538,
      "learning_rate": 0.00039000000000000005,
      "loss": 8.0054,
      "step": 130
    },
    {
      "epoch": 0.00131,
      "grad_norm": 0.781068809712636,
      "learning_rate": 0.000393,
      "loss": 7.9576,
      "step": 131
    },
    {
      "epoch": 0.00132,
      "grad_norm": 1.0367188815921278,
      "learning_rate": 0.00039600000000000003,
      "loss": 7.9591,
      "step": 132
    },
    {
      "epoch": 0.00133,
      "grad_norm": 0.7841067594214959,
      "learning_rate": 0.00039900000000000005,
      "loss": 7.9293,
      "step": 133
    },
    {
      "epoch": 0.00134,
      "grad_norm": 0.8594113418007541,
      "learning_rate": 0.000402,
      "loss": 7.9217,
      "step": 134
    },
    {
      "epoch": 0.00135,
      "grad_norm": 0.7468719464596628,
      "learning_rate": 0.00040500000000000003,
      "loss": 7.8883,
      "step": 135
    },
    {
      "epoch": 0.00136,
      "grad_norm": 0.8415432392576198,
      "learning_rate": 0.00040800000000000005,
      "loss": 7.8771,
      "step": 136
    },
    {
      "epoch": 0.00137,
      "grad_norm": 0.7386877288273068,
      "learning_rate": 0.000411,
      "loss": 7.8542,
      "step": 137
    },
    {
      "epoch": 0.00138,
      "grad_norm": 0.7450176106214967,
      "learning_rate": 0.00041400000000000003,
      "loss": 7.8322,
      "step": 138
    },
    {
      "epoch": 0.00139,
      "grad_norm": 0.7424698225823185,
      "learning_rate": 0.00041700000000000005,
      "loss": 7.8197,
      "step": 139
    },
    {
      "epoch": 0.0014,
      "grad_norm": 0.7268306219948636,
      "learning_rate": 0.00042000000000000007,
      "loss": 7.8018,
      "step": 140
    },
    {
      "epoch": 0.00141,
      "grad_norm": 0.7115037341291065,
      "learning_rate": 0.000423,
      "loss": 7.7905,
      "step": 141
    },
    {
      "epoch": 0.00142,
      "grad_norm": 0.6725409058271569,
      "learning_rate": 0.00042599999999999995,
      "loss": 7.7772,
      "step": 142
    },
    {
      "epoch": 0.00143,
      "grad_norm": 0.6830809074405504,
      "learning_rate": 0.00042899999999999997,
      "loss": 7.7496,
      "step": 143
    },
    {
      "epoch": 0.00144,
      "grad_norm": 0.6791849969278475,
      "learning_rate": 0.000432,
      "loss": 7.7318,
      "step": 144
    },
    {
      "epoch": 0.00145,
      "grad_norm": 0.6462817248800249,
      "learning_rate": 0.000435,
      "loss": 7.7251,
      "step": 145
    },
    {
      "epoch": 0.00146,
      "grad_norm": 0.6695930969912641,
      "learning_rate": 0.00043799999999999997,
      "loss": 7.6893,
      "step": 146
    },
    {
      "epoch": 0.00147,
      "grad_norm": 0.7012896651032599,
      "learning_rate": 0.000441,
      "loss": 7.6817,
      "step": 147
    },
    {
      "epoch": 0.00148,
      "grad_norm": 0.7755580877429182,
      "learning_rate": 0.000444,
      "loss": 7.6698,
      "step": 148
    },
    {
      "epoch": 0.00149,
      "grad_norm": 1.088947674236225,
      "learning_rate": 0.00044699999999999997,
      "loss": 7.6713,
      "step": 149
    },
    {
      "epoch": 0.0015,
      "grad_norm": 0.9396541473912592,
      "learning_rate": 0.00045,
      "loss": 7.6436,
      "step": 150
    },
    {
      "epoch": 0.00151,
      "grad_norm": 0.640217037835256,
      "learning_rate": 0.000453,
      "loss": 7.6133,
      "step": 151
    },
    {
      "epoch": 0.00152,
      "grad_norm": 0.9098496631236208,
      "learning_rate": 0.000456,
      "loss": 7.6149,
      "step": 152
    },
    {
      "epoch": 0.00153,
      "grad_norm": 0.6175759444520236,
      "learning_rate": 0.000459,
      "loss": 7.5962,
      "step": 153
    },
    {
      "epoch": 0.00154,
      "grad_norm": 0.6884616968083866,
      "learning_rate": 0.000462,
      "loss": 7.5772,
      "step": 154
    },
    {
      "epoch": 0.00155,
      "grad_norm": 0.6360938051960316,
      "learning_rate": 0.000465,
      "loss": 7.5598,
      "step": 155
    },
    {
      "epoch": 0.00156,
      "grad_norm": 0.6937352973699618,
      "learning_rate": 0.000468,
      "loss": 7.5366,
      "step": 156
    },
    {
      "epoch": 0.00157,
      "grad_norm": 0.6334556877551312,
      "learning_rate": 0.000471,
      "loss": 7.53,
      "step": 157
    },
    {
      "epoch": 0.00158,
      "grad_norm": 0.9707170671679272,
      "learning_rate": 0.00047400000000000003,
      "loss": 7.5153,
      "step": 158
    },
    {
      "epoch": 0.00159,
      "grad_norm": 1.0240750882994218,
      "learning_rate": 0.000477,
      "loss": 7.5076,
      "step": 159
    },
    {
      "epoch": 0.0016,
      "grad_norm": 0.9322894026235434,
      "learning_rate": 0.00048,
      "loss": 7.4836,
      "step": 160
    },
    {
      "epoch": 0.00161,
      "grad_norm": 0.5279037898518898,
      "learning_rate": 0.00048300000000000003,
      "loss": 7.4527,
      "step": 161
    },
    {
      "epoch": 0.00162,
      "grad_norm": 0.687356662308957,
      "learning_rate": 0.00048600000000000005,
      "loss": 7.4601,
      "step": 162
    },
    {
      "epoch": 0.00163,
      "grad_norm": 0.5623951705594973,
      "learning_rate": 0.0004890000000000001,
      "loss": 7.4388,
      "step": 163
    },
    {
      "epoch": 0.00164,
      "grad_norm": 0.5581337114560441,
      "learning_rate": 0.000492,
      "loss": 7.4399,
      "step": 164
    },
    {
      "epoch": 0.00165,
      "grad_norm": 0.5516159301488641,
      "learning_rate": 0.000495,
      "loss": 7.4126,
      "step": 165
    },
    {
      "epoch": 0.00166,
      "grad_norm": 0.5242244578051735,
      "learning_rate": 0.0004980000000000001,
      "loss": 7.3876,
      "step": 166
    },
    {
      "epoch": 0.00167,
      "grad_norm": 0.45997959649003123,
      "learning_rate": 0.000501,
      "loss": 7.3779,
      "step": 167
    },
    {
      "epoch": 0.00168,
      "grad_norm": 0.5436289820614866,
      "learning_rate": 0.000504,
      "loss": 7.3569,
      "step": 168
    },
    {
      "epoch": 0.00169,
      "grad_norm": 0.4983067598465849,
      "learning_rate": 0.0005070000000000001,
      "loss": 7.3495,
      "step": 169
    },
    {
      "epoch": 0.0017,
      "grad_norm": 0.4402852485923817,
      "learning_rate": 0.00051,
      "loss": 7.3316,
      "step": 170
    },
    {
      "epoch": 0.00171,
      "grad_norm": 0.5221521396945228,
      "learning_rate": 0.000513,
      "loss": 7.3138,
      "step": 171
    },
    {
      "epoch": 0.00172,
      "grad_norm": 0.45544347662440743,
      "learning_rate": 0.000516,
      "loss": 7.3129,
      "step": 172
    },
    {
      "epoch": 0.00173,
      "grad_norm": 0.4745602833877857,
      "learning_rate": 0.0005189999999999999,
      "loss": 7.2961,
      "step": 173
    },
    {
      "epoch": 0.00174,
      "grad_norm": 0.5121580158942174,
      "learning_rate": 0.000522,
      "loss": 7.311,
      "step": 174
    },
    {
      "epoch": 0.00175,
      "grad_norm": 0.680505499537256,
      "learning_rate": 0.000525,
      "loss": 7.2769,
      "step": 175
    },
    {
      "epoch": 0.00176,
      "grad_norm": 0.7210959926983863,
      "learning_rate": 0.0005279999999999999,
      "loss": 7.2549,
      "step": 176
    },
    {
      "epoch": 0.00177,
      "grad_norm": 0.7510526045152774,
      "learning_rate": 0.000531,
      "loss": 7.2572,
      "step": 177
    },
    {
      "epoch": 0.00178,
      "grad_norm": 0.7343359791017195,
      "learning_rate": 0.000534,
      "loss": 7.2506,
      "step": 178
    },
    {
      "epoch": 0.00179,
      "grad_norm": 0.6185909230661502,
      "learning_rate": 0.000537,
      "loss": 7.2371,
      "step": 179
    },
    {
      "epoch": 0.0018,
      "grad_norm": 0.42887581745789505,
      "learning_rate": 0.00054,
      "loss": 7.2041,
      "step": 180
    },
    {
      "epoch": 0.00181,
      "grad_norm": 0.5359194273155663,
      "learning_rate": 0.000543,
      "loss": 7.1961,
      "step": 181
    },
    {
      "epoch": 0.00182,
      "grad_norm": 0.4369454993609359,
      "learning_rate": 0.000546,
      "loss": 7.188,
      "step": 182
    },
    {
      "epoch": 0.00183,
      "grad_norm": 0.41001210068633426,
      "learning_rate": 0.000549,
      "loss": 7.1769,
      "step": 183
    },
    {
      "epoch": 0.00184,
      "grad_norm": 0.4396906853721559,
      "learning_rate": 0.000552,
      "loss": 7.1646,
      "step": 184
    },
    {
      "epoch": 0.00185,
      "grad_norm": 0.4311329351383538,
      "learning_rate": 0.000555,
      "loss": 7.1403,
      "step": 185
    },
    {
      "epoch": 0.00186,
      "grad_norm": 0.44013221474943204,
      "learning_rate": 0.000558,
      "loss": 7.1495,
      "step": 186
    },
    {
      "epoch": 0.00187,
      "grad_norm": 0.42719641993731927,
      "learning_rate": 0.000561,
      "loss": 7.1244,
      "step": 187
    },
    {
      "epoch": 0.00188,
      "grad_norm": 0.39192004433941763,
      "learning_rate": 0.000564,
      "loss": 7.1022,
      "step": 188
    },
    {
      "epoch": 0.00189,
      "grad_norm": 0.5474017259570552,
      "learning_rate": 0.000567,
      "loss": 7.1002,
      "step": 189
    },
    {
      "epoch": 0.0019,
      "grad_norm": 0.5320208949420774,
      "learning_rate": 0.00057,
      "loss": 7.0824,
      "step": 190
    },
    {
      "epoch": 0.00191,
      "grad_norm": 0.7661733453917681,
      "learning_rate": 0.000573,
      "loss": 7.0764,
      "step": 191
    },
    {
      "epoch": 0.00192,
      "grad_norm": 0.9858385620454592,
      "learning_rate": 0.000576,
      "loss": 7.0775,
      "step": 192
    },
    {
      "epoch": 0.00193,
      "grad_norm": 0.9591489873604585,
      "learning_rate": 0.000579,
      "loss": 7.0725,
      "step": 193
    },
    {
      "epoch": 0.00194,
      "grad_norm": 0.8774582305545432,
      "learning_rate": 0.000582,
      "loss": 7.0715,
      "step": 194
    },
    {
      "epoch": 0.00195,
      "grad_norm": 0.9015156284206135,
      "learning_rate": 0.000585,
      "loss": 7.0487,
      "step": 195
    },
    {
      "epoch": 0.00196,
      "grad_norm": 0.8000421121121074,
      "learning_rate": 0.000588,
      "loss": 7.0339,
      "step": 196
    },
    {
      "epoch": 0.00197,
      "grad_norm": 0.8468645119231468,
      "learning_rate": 0.000591,
      "loss": 7.0329,
      "step": 197
    },
    {
      "epoch": 0.00198,
      "grad_norm": 0.6924161058762034,
      "learning_rate": 0.000594,
      "loss": 7.0197,
      "step": 198
    },
    {
      "epoch": 0.00199,
      "grad_norm": 0.5671884633245193,
      "learning_rate": 0.0005970000000000001,
      "loss": 6.9992,
      "step": 199
    },
    {
      "epoch": 0.002,
      "grad_norm": 0.5557515599131739,
      "learning_rate": 0.0006000000000000001,
      "loss": 6.9982,
      "step": 200
    },
    {
      "epoch": 0.00201,
      "grad_norm": 0.4619645912916116,
      "learning_rate": 0.000603,
      "loss": 6.9694,
      "step": 201
    },
    {
      "epoch": 0.00202,
      "grad_norm": 0.4929222182059183,
      "learning_rate": 0.0006060000000000001,
      "loss": 6.9601,
      "step": 202
    },
    {
      "epoch": 0.00203,
      "grad_norm": 0.37702087689736097,
      "learning_rate": 0.0006090000000000001,
      "loss": 6.9599,
      "step": 203
    },
    {
      "epoch": 0.00204,
      "grad_norm": 0.45518892473332057,
      "learning_rate": 0.000612,
      "loss": 6.9476,
      "step": 204
    },
    {
      "epoch": 0.00205,
      "grad_norm": 0.384809834745988,
      "learning_rate": 0.000615,
      "loss": 6.9322,
      "step": 205
    },
    {
      "epoch": 0.00206,
      "grad_norm": 0.38976835520829006,
      "learning_rate": 0.000618,
      "loss": 6.9255,
      "step": 206
    },
    {
      "epoch": 0.00207,
      "grad_norm": 0.4375749283114229,
      "learning_rate": 0.000621,
      "loss": 6.9166,
      "step": 207
    },
    {
      "epoch": 0.00208,
      "grad_norm": 0.6317927736115376,
      "learning_rate": 0.000624,
      "loss": 6.9157,
      "step": 208
    },
    {
      "epoch": 0.00209,
      "grad_norm": 0.7772617938427908,
      "learning_rate": 0.000627,
      "loss": 6.9007,
      "step": 209
    },
    {
      "epoch": 0.0021,
      "grad_norm": 1.2298132092555871,
      "learning_rate": 0.00063,
      "loss": 6.9015,
      "step": 210
    },
    {
      "epoch": 0.00211,
      "grad_norm": 0.7435502579532725,
      "learning_rate": 0.000633,
      "loss": 6.8883,
      "step": 211
    },
    {
      "epoch": 0.00212,
      "grad_norm": 0.3794792840193541,
      "learning_rate": 0.000636,
      "loss": 6.8693,
      "step": 212
    },
    {
      "epoch": 0.00213,
      "grad_norm": 0.688114698891817,
      "learning_rate": 0.000639,
      "loss": 6.8589,
      "step": 213
    },
    {
      "epoch": 0.00214,
      "grad_norm": 0.49868168831557785,
      "learning_rate": 0.000642,
      "loss": 6.8504,
      "step": 214
    },
    {
      "epoch": 0.00215,
      "grad_norm": 0.6292678113208914,
      "learning_rate": 0.000645,
      "loss": 6.8506,
      "step": 215
    },
    {
      "epoch": 0.00216,
      "grad_norm": 0.3673676300147008,
      "learning_rate": 0.000648,
      "loss": 6.8383,
      "step": 216
    },
    {
      "epoch": 0.00217,
      "grad_norm": 0.4511278951279821,
      "learning_rate": 0.000651,
      "loss": 6.8131,
      "step": 217
    },
    {
      "epoch": 0.00218,
      "grad_norm": 0.3783399316849399,
      "learning_rate": 0.000654,
      "loss": 6.8146,
      "step": 218
    },
    {
      "epoch": 0.00219,
      "grad_norm": 0.3651267362185199,
      "learning_rate": 0.000657,
      "loss": 6.7867,
      "step": 219
    },
    {
      "epoch": 0.0022,
      "grad_norm": 0.4708567379601776,
      "learning_rate": 0.00066,
      "loss": 6.8074,
      "step": 220
    },
    {
      "epoch": 0.00221,
      "grad_norm": 0.5218809374007617,
      "learning_rate": 0.0006630000000000001,
      "loss": 6.7837,
      "step": 221
    },
    {
      "epoch": 0.00222,
      "grad_norm": 0.7518789100021657,
      "learning_rate": 0.000666,
      "loss": 6.7774,
      "step": 222
    },
    {
      "epoch": 0.00223,
      "grad_norm": 0.9964198530393009,
      "learning_rate": 0.000669,
      "loss": 6.7824,
      "step": 223
    },
    {
      "epoch": 0.00224,
      "grad_norm": 0.807522133159797,
      "learning_rate": 0.0006720000000000001,
      "loss": 6.7748,
      "step": 224
    },
    {
      "epoch": 0.00225,
      "grad_norm": 0.6803785939854445,
      "learning_rate": 0.000675,
      "loss": 6.7563,
      "step": 225
    },
    {
      "epoch": 0.00226,
      "grad_norm": 1.0762485305388094,
      "learning_rate": 0.000678,
      "loss": 6.7681,
      "step": 226
    },
    {
      "epoch": 0.00227,
      "grad_norm": 0.9129777590140102,
      "learning_rate": 0.0006810000000000001,
      "loss": 6.734,
      "step": 227
    },
    {
      "epoch": 0.00228,
      "grad_norm": 1.299087365220233,
      "learning_rate": 0.000684,
      "loss": 6.7517,
      "step": 228
    },
    {
      "epoch": 0.00229,
      "grad_norm": 0.7392729796935101,
      "learning_rate": 0.000687,
      "loss": 6.7331,
      "step": 229
    },
    {
      "epoch": 0.0023,
      "grad_norm": 0.6158971034672628,
      "learning_rate": 0.0006900000000000001,
      "loss": 6.7089,
      "step": 230
    },
    {
      "epoch": 0.00231,
      "grad_norm": 0.7016695647497186,
      "learning_rate": 0.000693,
      "loss": 6.6956,
      "step": 231
    },
    {
      "epoch": 0.00232,
      "grad_norm": 0.6434185638703606,
      "learning_rate": 0.000696,
      "loss": 6.7082,
      "step": 232
    },
    {
      "epoch": 0.00233,
      "grad_norm": 0.6044879418578446,
      "learning_rate": 0.0006990000000000001,
      "loss": 6.6921,
      "step": 233
    },
    {
      "epoch": 0.00234,
      "grad_norm": 0.4480061456613071,
      "learning_rate": 0.000702,
      "loss": 6.679,
      "step": 234
    },
    {
      "epoch": 0.00235,
      "grad_norm": 0.4492075259819563,
      "learning_rate": 0.000705,
      "loss": 6.6661,
      "step": 235
    },
    {
      "epoch": 0.00236,
      "grad_norm": 0.3889271878969786,
      "learning_rate": 0.000708,
      "loss": 6.6458,
      "step": 236
    },
    {
      "epoch": 0.00237,
      "grad_norm": 0.41607485698419117,
      "learning_rate": 0.0007109999999999999,
      "loss": 6.6575,
      "step": 237
    },
    {
      "epoch": 0.00238,
      "grad_norm": 0.3515958046168432,
      "learning_rate": 0.000714,
      "loss": 6.6432,
      "step": 238
    },
    {
      "epoch": 0.00239,
      "grad_norm": 0.4049983983930021,
      "learning_rate": 0.000717,
      "loss": 6.6274,
      "step": 239
    },
    {
      "epoch": 0.0024,
      "grad_norm": 0.3247748587680522,
      "learning_rate": 0.0007199999999999999,
      "loss": 6.6414,
      "step": 240
    },
    {
      "epoch": 0.00241,
      "grad_norm": 0.36262136655648425,
      "learning_rate": 0.000723,
      "loss": 6.6267,
      "step": 241
    },
    {
      "epoch": 0.00242,
      "grad_norm": 0.3238608381517686,
      "learning_rate": 0.000726,
      "loss": 6.598,
      "step": 242
    },
    {
      "epoch": 0.00243,
      "grad_norm": 0.31681258130419926,
      "learning_rate": 0.000729,
      "loss": 6.5877,
      "step": 243
    },
    {
      "epoch": 0.00244,
      "grad_norm": 0.3616370322868285,
      "learning_rate": 0.000732,
      "loss": 6.5911,
      "step": 244
    },
    {
      "epoch": 0.00245,
      "grad_norm": 0.4634076592576489,
      "learning_rate": 0.000735,
      "loss": 6.5815,
      "step": 245
    },
    {
      "epoch": 0.00246,
      "grad_norm": 0.7486687333799987,
      "learning_rate": 0.000738,
      "loss": 6.5693,
      "step": 246
    },
    {
      "epoch": 0.00247,
      "grad_norm": 1.662590620725261,
      "learning_rate": 0.000741,
      "loss": 6.5942,
      "step": 247
    },
    {
      "epoch": 0.00248,
      "grad_norm": 0.7962452736060305,
      "learning_rate": 0.000744,
      "loss": 6.5799,
      "step": 248
    },
    {
      "epoch": 0.00249,
      "grad_norm": 1.091558045553116,
      "learning_rate": 0.000747,
      "loss": 6.5681,
      "step": 249
    },
    {
      "epoch": 0.0025,
      "grad_norm": 1.1566118279124307,
      "learning_rate": 0.00075,
      "loss": 6.5821,
      "step": 250
    },
    {
      "epoch": 0.00251,
      "grad_norm": 0.7784764288643516,
      "learning_rate": 0.000753,
      "loss": 6.5311,
      "step": 251
    },
    {
      "epoch": 0.00252,
      "grad_norm": 0.8327004278614876,
      "learning_rate": 0.000756,
      "loss": 6.5476,
      "step": 252
    },
    {
      "epoch": 0.00253,
      "grad_norm": 0.8184368732684698,
      "learning_rate": 0.000759,
      "loss": 6.5327,
      "step": 253
    },
    {
      "epoch": 0.00254,
      "grad_norm": 1.1486609061260633,
      "learning_rate": 0.000762,
      "loss": 6.541,
      "step": 254
    },
    {
      "epoch": 0.00255,
      "grad_norm": 0.8085196503687528,
      "learning_rate": 0.0007650000000000001,
      "loss": 6.533,
      "step": 255
    },
    {
      "epoch": 0.00256,
      "grad_norm": 0.6227788989369496,
      "learning_rate": 0.000768,
      "loss": 6.5039,
      "step": 256
    },
    {
      "epoch": 0.00257,
      "grad_norm": 0.7880922816831363,
      "learning_rate": 0.000771,
      "loss": 6.5163,
      "step": 257
    },
    {
      "epoch": 0.00258,
      "grad_norm": 0.8888953354311911,
      "learning_rate": 0.0007740000000000001,
      "loss": 6.497,
      "step": 258
    },
    {
      "epoch": 0.00259,
      "grad_norm": 0.6556759438280303,
      "learning_rate": 0.000777,
      "loss": 6.491,
      "step": 259
    },
    {
      "epoch": 0.0026,
      "grad_norm": 0.5556045743581063,
      "learning_rate": 0.0007800000000000001,
      "loss": 6.4736,
      "step": 260
    },
    {
      "epoch": 0.00261,
      "grad_norm": 0.5757853625150707,
      "learning_rate": 0.0007830000000000001,
      "loss": 6.4835,
      "step": 261
    },
    {
      "epoch": 0.00262,
      "grad_norm": 0.4702991355160636,
      "learning_rate": 0.000786,
      "loss": 6.4605,
      "step": 262
    },
    {
      "epoch": 0.00263,
      "grad_norm": 0.5236270398277312,
      "learning_rate": 0.0007890000000000001,
      "loss": 6.4668,
      "step": 263
    },
    {
      "epoch": 0.00264,
      "grad_norm": 0.57918798311576,
      "learning_rate": 0.0007920000000000001,
      "loss": 6.4494,
      "step": 264
    },
    {
      "epoch": 0.00265,
      "grad_norm": 0.48386144230733374,
      "learning_rate": 0.000795,
      "loss": 6.4394,
      "step": 265
    },
    {
      "epoch": 0.00266,
      "grad_norm": 0.44170660781498655,
      "learning_rate": 0.0007980000000000001,
      "loss": 6.4372,
      "step": 266
    },
    {
      "epoch": 0.00267,
      "grad_norm": 0.35002993788441544,
      "learning_rate": 0.0008010000000000001,
      "loss": 6.4273,
      "step": 267
    },
    {
      "epoch": 0.00268,
      "grad_norm": 0.49450621569184094,
      "learning_rate": 0.000804,
      "loss": 6.426,
      "step": 268
    },
    {
      "epoch": 0.00269,
      "grad_norm": 0.5917094821390496,
      "learning_rate": 0.0008070000000000001,
      "loss": 6.4104,
      "step": 269
    },
    {
      "epoch": 0.0027,
      "grad_norm": 0.8610573273382283,
      "learning_rate": 0.0008100000000000001,
      "loss": 6.4165,
      "step": 270
    },
    {
      "epoch": 0.00271,
      "grad_norm": 0.9687752227819599,
      "learning_rate": 0.000813,
      "loss": 6.4342,
      "step": 271
    },
    {
      "epoch": 0.00272,
      "grad_norm": 0.8252297813091672,
      "learning_rate": 0.0008160000000000001,
      "loss": 6.4109,
      "step": 272
    },
    {
      "epoch": 0.00273,
      "grad_norm": 1.0509340869742279,
      "learning_rate": 0.0008190000000000001,
      "loss": 6.4091,
      "step": 273
    },
    {
      "epoch": 0.00274,
      "grad_norm": 1.2492890299831212,
      "learning_rate": 0.000822,
      "loss": 6.3984,
      "step": 274
    },
    {
      "epoch": 0.00275,
      "grad_norm": 0.7449833128353471,
      "learning_rate": 0.0008250000000000001,
      "loss": 6.3937,
      "step": 275
    },
    {
      "epoch": 0.00276,
      "grad_norm": 0.650217944416595,
      "learning_rate": 0.0008280000000000001,
      "loss": 6.383,
      "step": 276
    },
    {
      "epoch": 0.00277,
      "grad_norm": 0.6897831829588578,
      "learning_rate": 0.0008310000000000001,
      "loss": 6.3791,
      "step": 277
    },
    {
      "epoch": 0.00278,
      "grad_norm": 0.6845751052836879,
      "learning_rate": 0.0008340000000000001,
      "loss": 6.3625,
      "step": 278
    },
    {
      "epoch": 0.00279,
      "grad_norm": 0.6895680336476154,
      "learning_rate": 0.0008370000000000001,
      "loss": 6.3607,
      "step": 279
    },
    {
      "epoch": 0.0028,
      "grad_norm": 0.7962805776833526,
      "learning_rate": 0.0008400000000000001,
      "loss": 6.3497,
      "step": 280
    },
    {
      "epoch": 0.00281,
      "grad_norm": 0.7281805890288613,
      "learning_rate": 0.0008430000000000001,
      "loss": 6.3358,
      "step": 281
    },
    {
      "epoch": 0.00282,
      "grad_norm": 0.6077351466654652,
      "learning_rate": 0.000846,
      "loss": 6.3343,
      "step": 282
    },
    {
      "epoch": 0.00283,
      "grad_norm": 0.7021452115418912,
      "learning_rate": 0.0008489999999999999,
      "loss": 6.3268,
      "step": 283
    },
    {
      "epoch": 0.00284,
      "grad_norm": 0.8067589645945545,
      "learning_rate": 0.0008519999999999999,
      "loss": 6.3217,
      "step": 284
    },
    {
      "epoch": 0.00285,
      "grad_norm": 0.7238661685202905,
      "learning_rate": 0.000855,
      "loss": 6.3229,
      "step": 285
    },
    {
      "epoch": 0.00286,
      "grad_norm": 0.6566110557704579,
      "learning_rate": 0.0008579999999999999,
      "loss": 6.3002,
      "step": 286
    },
    {
      "epoch": 0.00287,
      "grad_norm": 0.5448339024770075,
      "learning_rate": 0.000861,
      "loss": 6.3016,
      "step": 287
    },
    {
      "epoch": 0.00288,
      "grad_norm": 0.4516947530472597,
      "learning_rate": 0.000864,
      "loss": 6.279,
      "step": 288
    },
    {
      "epoch": 0.00289,
      "grad_norm": 0.5415480649305122,
      "learning_rate": 0.0008669999999999999,
      "loss": 6.2929,
      "step": 289
    },
    {
      "epoch": 0.0029,
      "grad_norm": 0.5812113854990302,
      "learning_rate": 0.00087,
      "loss": 6.2743,
      "step": 290
    },
    {
      "epoch": 0.00291,
      "grad_norm": 0.5620937871349196,
      "learning_rate": 0.000873,
      "loss": 6.2679,
      "step": 291
    },
    {
      "epoch": 0.00292,
      "grad_norm": 0.43362286284353735,
      "learning_rate": 0.0008759999999999999,
      "loss": 6.2618,
      "step": 292
    },
    {
      "epoch": 0.00293,
      "grad_norm": 0.563953479113573,
      "learning_rate": 0.000879,
      "loss": 6.2491,
      "step": 293
    },
    {
      "epoch": 0.00294,
      "grad_norm": 0.5784107789505917,
      "learning_rate": 0.000882,
      "loss": 6.2636,
      "step": 294
    },
    {
      "epoch": 0.00295,
      "grad_norm": 0.5410292555611181,
      "learning_rate": 0.0008849999999999999,
      "loss": 6.2293,
      "step": 295
    },
    {
      "epoch": 0.00296,
      "grad_norm": 0.5413778994680785,
      "learning_rate": 0.000888,
      "loss": 6.2428,
      "step": 296
    },
    {
      "epoch": 0.00297,
      "grad_norm": 0.7587824592105622,
      "learning_rate": 0.000891,
      "loss": 6.2186,
      "step": 297
    },
    {
      "epoch": 0.00298,
      "grad_norm": 1.2390223277612324,
      "learning_rate": 0.0008939999999999999,
      "loss": 6.2454,
      "step": 298
    },
    {
      "epoch": 0.00299,
      "grad_norm": 1.0051162436627608,
      "learning_rate": 0.000897,
      "loss": 6.2195,
      "step": 299
    },
    {
      "epoch": 0.003,
      "grad_norm": 0.9301924220055711,
      "learning_rate": 0.0009,
      "loss": 6.2223,
      "step": 300
    },
    {
      "epoch": 0.00301,
      "grad_norm": 0.9092669241988458,
      "learning_rate": 0.0009029999999999999,
      "loss": 6.2133,
      "step": 301
    },
    {
      "epoch": 0.00302,
      "grad_norm": 1.0280476917286576,
      "learning_rate": 0.000906,
      "loss": 6.2045,
      "step": 302
    },
    {
      "epoch": 0.00303,
      "grad_norm": 0.955686909103224,
      "learning_rate": 0.000909,
      "loss": 6.2159,
      "step": 303
    },
    {
      "epoch": 0.00304,
      "grad_norm": 0.924153033838841,
      "learning_rate": 0.000912,
      "loss": 6.2068,
      "step": 304
    },
    {
      "epoch": 0.00305,
      "grad_norm": 1.0129569357421315,
      "learning_rate": 0.000915,
      "loss": 6.2162,
      "step": 305
    },
    {
      "epoch": 0.00306,
      "grad_norm": 0.908196615580651,
      "learning_rate": 0.000918,
      "loss": 6.1982,
      "step": 306
    },
    {
      "epoch": 0.00307,
      "grad_norm": 0.9412115799039344,
      "learning_rate": 0.000921,
      "loss": 6.1922,
      "step": 307
    },
    {
      "epoch": 0.00308,
      "grad_norm": 1.0133210251014897,
      "learning_rate": 0.000924,
      "loss": 6.1801,
      "step": 308
    },
    {
      "epoch": 0.00309,
      "grad_norm": 0.7885110954474147,
      "learning_rate": 0.000927,
      "loss": 6.1916,
      "step": 309
    },
    {
      "epoch": 0.0031,
      "grad_norm": 0.8502930513548597,
      "learning_rate": 0.00093,
      "loss": 6.1767,
      "step": 310
    },
    {
      "epoch": 0.00311,
      "grad_norm": 0.8830498260903274,
      "learning_rate": 0.000933,
      "loss": 6.1706,
      "step": 311
    },
    {
      "epoch": 0.00312,
      "grad_norm": 0.7731682808518218,
      "learning_rate": 0.000936,
      "loss": 6.1483,
      "step": 312
    },
    {
      "epoch": 0.00313,
      "grad_norm": 0.5034693382195244,
      "learning_rate": 0.0009390000000000001,
      "loss": 6.1541,
      "step": 313
    },
    {
      "epoch": 0.00314,
      "grad_norm": 0.5944206771214186,
      "learning_rate": 0.000942,
      "loss": 6.1615,
      "step": 314
    },
    {
      "epoch": 0.00315,
      "grad_norm": 0.6198949865082991,
      "learning_rate": 0.000945,
      "loss": 6.1382,
      "step": 315
    },
    {
      "epoch": 0.00316,
      "grad_norm": 0.6322301787116348,
      "learning_rate": 0.0009480000000000001,
      "loss": 6.112,
      "step": 316
    },
    {
      "epoch": 0.00317,
      "grad_norm": 0.6034799354522895,
      "learning_rate": 0.000951,
      "loss": 6.1294,
      "step": 317
    },
    {
      "epoch": 0.00318,
      "grad_norm": 0.594812791501401,
      "learning_rate": 0.000954,
      "loss": 6.1243,
      "step": 318
    },
    {
      "epoch": 0.00319,
      "grad_norm": 0.5121880287121718,
      "learning_rate": 0.0009570000000000001,
      "loss": 6.1104,
      "step": 319
    },
    {
      "epoch": 0.0032,
      "grad_norm": 0.4383869814323231,
      "learning_rate": 0.00096,
      "loss": 6.1078,
      "step": 320
    },
    {
      "epoch": 0.00321,
      "grad_norm": 0.46253996577569373,
      "learning_rate": 0.000963,
      "loss": 6.0957,
      "step": 321
    },
    {
      "epoch": 0.00322,
      "grad_norm": 0.5330974668531825,
      "learning_rate": 0.0009660000000000001,
      "loss": 6.0852,
      "step": 322
    },
    {
      "epoch": 0.00323,
      "grad_norm": 0.4269664457046623,
      "learning_rate": 0.000969,
      "loss": 6.0805,
      "step": 323
    },
    {
      "epoch": 0.00324,
      "grad_norm": 0.4120363869720864,
      "learning_rate": 0.0009720000000000001,
      "loss": 6.0833,
      "step": 324
    },
    {
      "epoch": 0.00325,
      "grad_norm": 0.43852524771556284,
      "learning_rate": 0.0009750000000000001,
      "loss": 6.0494,
      "step": 325
    },
    {
      "epoch": 0.00326,
      "grad_norm": 0.5390182954459198,
      "learning_rate": 0.0009780000000000001,
      "loss": 6.0675,
      "step": 326
    },
    {
      "epoch": 0.00327,
      "grad_norm": 0.7655676469366879,
      "learning_rate": 0.000981,
      "loss": 6.0729,
      "step": 327
    },
    {
      "epoch": 0.00328,
      "grad_norm": 1.3095809996152918,
      "learning_rate": 0.000984,
      "loss": 6.0661,
      "step": 328
    },
    {
      "epoch": 0.00329,
      "grad_norm": 0.9398963728555475,
      "learning_rate": 0.000987,
      "loss": 6.0593,
      "step": 329
    },
    {
      "epoch": 0.0033,
      "grad_norm": 0.9687365266328093,
      "learning_rate": 0.00099,
      "loss": 6.0595,
      "step": 330
    },
    {
      "epoch": 0.00331,
      "grad_norm": 1.0572725295338647,
      "learning_rate": 0.0009930000000000002,
      "loss": 6.0709,
      "step": 331
    },
    {
      "epoch": 0.00332,
      "grad_norm": 0.7804253243852415,
      "learning_rate": 0.0009960000000000001,
      "loss": 6.0449,
      "step": 332
    },
    {
      "epoch": 0.00333,
      "grad_norm": 0.9477444786898163,
      "learning_rate": 0.000999,
      "loss": 6.0352,
      "step": 333
    },
    {
      "epoch": 0.00334,
      "grad_norm": 1.119636024931192,
      "learning_rate": 0.001002,
      "loss": 6.0509,
      "step": 334
    },
    {
      "epoch": 0.00335,
      "grad_norm": 1.0398096390896505,
      "learning_rate": 0.001005,
      "loss": 6.0495,
      "step": 335
    },
    {
      "epoch": 0.00336,
      "grad_norm": 0.8946811171098662,
      "learning_rate": 0.001008,
      "loss": 6.0348,
      "step": 336
    },
    {
      "epoch": 0.00337,
      "grad_norm": 1.0240126692017184,
      "learning_rate": 0.0010110000000000002,
      "loss": 6.0215,
      "step": 337
    },
    {
      "epoch": 0.00338,
      "grad_norm": 1.0245564805630587,
      "learning_rate": 0.0010140000000000001,
      "loss": 6.0414,
      "step": 338
    },
    {
      "epoch": 0.00339,
      "grad_norm": 0.9606863659347135,
      "learning_rate": 0.0010170000000000001,
      "loss": 6.0439,
      "step": 339
    },
    {
      "epoch": 0.0034,
      "grad_norm": 0.7549695201896788,
      "learning_rate": 0.00102,
      "loss": 6.0067,
      "step": 340
    },
    {
      "epoch": 0.00341,
      "grad_norm": 0.7981881225838676,
      "learning_rate": 0.001023,
      "loss": 5.9935,
      "step": 341
    },
    {
      "epoch": 0.00342,
      "grad_norm": 0.6476806075689499,
      "learning_rate": 0.001026,
      "loss": 5.9919,
      "step": 342
    },
    {
      "epoch": 0.00343,
      "grad_norm": 0.6180530813345404,
      "learning_rate": 0.0010290000000000002,
      "loss": 5.9943,
      "step": 343
    },
    {
      "epoch": 0.00344,
      "grad_norm": 0.5904235085071222,
      "learning_rate": 0.001032,
      "loss": 5.9824,
      "step": 344
    },
    {
      "epoch": 0.00345,
      "grad_norm": 0.5662246103658737,
      "learning_rate": 0.001035,
      "loss": 5.9736,
      "step": 345
    },
    {
      "epoch": 0.00346,
      "grad_norm": 0.5815647759401746,
      "learning_rate": 0.0010379999999999999,
      "loss": 5.9787,
      "step": 346
    },
    {
      "epoch": 0.00347,
      "grad_norm": 0.5922585456492798,
      "learning_rate": 0.001041,
      "loss": 5.9505,
      "step": 347
    },
    {
      "epoch": 0.00348,
      "grad_norm": 0.546770754957902,
      "learning_rate": 0.001044,
      "loss": 5.9566,
      "step": 348
    },
    {
      "epoch": 0.00349,
      "grad_norm": 0.5259270224752429,
      "learning_rate": 0.001047,
      "loss": 5.9388,
      "step": 349
    },
    {
      "epoch": 0.0035,
      "grad_norm": 0.5639453872626701,
      "learning_rate": 0.00105,
      "loss": 5.96,
      "step": 350
    },
    {
      "epoch": 0.00351,
      "grad_norm": 0.562716119705,
      "learning_rate": 0.001053,
      "loss": 5.9575,
      "step": 351
    },
    {
      "epoch": 0.00352,
      "grad_norm": 0.6276157353012132,
      "learning_rate": 0.0010559999999999999,
      "loss": 5.9357,
      "step": 352
    },
    {
      "epoch": 0.00353,
      "grad_norm": 0.6899105072430818,
      "learning_rate": 0.001059,
      "loss": 5.939,
      "step": 353
    },
    {
      "epoch": 0.00354,
      "grad_norm": 0.7414855002893123,
      "learning_rate": 0.001062,
      "loss": 5.9142,
      "step": 354
    },
    {
      "epoch": 0.00355,
      "grad_norm": 0.8275245559939801,
      "learning_rate": 0.001065,
      "loss": 5.932,
      "step": 355
    },
    {
      "epoch": 0.00356,
      "grad_norm": 0.8334699202442165,
      "learning_rate": 0.001068,
      "loss": 5.925,
      "step": 356
    },
    {
      "epoch": 0.00357,
      "grad_norm": 0.733463396370079,
      "learning_rate": 0.001071,
      "loss": 5.9147,
      "step": 357
    },
    {
      "epoch": 0.00358,
      "grad_norm": 0.5898600387972718,
      "learning_rate": 0.001074,
      "loss": 5.9022,
      "step": 358
    },
    {
      "epoch": 0.00359,
      "grad_norm": 0.53807120274405,
      "learning_rate": 0.001077,
      "loss": 5.8933,
      "step": 359
    },
    {
      "epoch": 0.0036,
      "grad_norm": 0.5868526923580266,
      "learning_rate": 0.00108,
      "loss": 5.8984,
      "step": 360
    },
    {
      "epoch": 0.00361,
      "grad_norm": 0.6561371025439784,
      "learning_rate": 0.001083,
      "loss": 5.8927,
      "step": 361
    },
    {
      "epoch": 0.00362,
      "grad_norm": 1.003392418604553,
      "learning_rate": 0.001086,
      "loss": 5.8918,
      "step": 362
    },
    {
      "epoch": 0.00363,
      "grad_norm": 1.2318504963643775,
      "learning_rate": 0.001089,
      "loss": 5.899,
      "step": 363
    },
    {
      "epoch": 0.00364,
      "grad_norm": 0.7073595215993886,
      "learning_rate": 0.001092,
      "loss": 5.8956,
      "step": 364
    },
    {
      "epoch": 0.00365,
      "grad_norm": 0.6834408409212124,
      "learning_rate": 0.001095,
      "loss": 5.8924,
      "step": 365
    },
    {
      "epoch": 0.00366,
      "grad_norm": 0.808409762735137,
      "learning_rate": 0.001098,
      "loss": 5.8732,
      "step": 366
    },
    {
      "epoch": 0.00367,
      "grad_norm": 1.150363353399194,
      "learning_rate": 0.001101,
      "loss": 5.8871,
      "step": 367
    },
    {
      "epoch": 0.00368,
      "grad_norm": 1.5759048240116487,
      "learning_rate": 0.001104,
      "loss": 5.9113,
      "step": 368
    },
    {
      "epoch": 0.00369,
      "grad_norm": 0.6706600035773841,
      "learning_rate": 0.001107,
      "loss": 5.8826,
      "step": 369
    },
    {
      "epoch": 0.0037,
      "grad_norm": 1.0165372867603926,
      "learning_rate": 0.00111,
      "loss": 5.8773,
      "step": 370
    },
    {
      "epoch": 0.00371,
      "grad_norm": 1.199271508025637,
      "learning_rate": 0.001113,
      "loss": 5.8595,
      "step": 371
    },
    {
      "epoch": 0.00372,
      "grad_norm": 0.8890339964058547,
      "learning_rate": 0.001116,
      "loss": 5.8693,
      "step": 372
    },
    {
      "epoch": 0.00373,
      "grad_norm": 1.0853395614111072,
      "learning_rate": 0.001119,
      "loss": 5.8674,
      "step": 373
    },
    {
      "epoch": 0.00374,
      "grad_norm": 0.9530307319001724,
      "learning_rate": 0.001122,
      "loss": 5.8703,
      "step": 374
    },
    {
      "epoch": 0.00375,
      "grad_norm": 0.934750801302772,
      "learning_rate": 0.0011250000000000001,
      "loss": 5.8719,
      "step": 375
    },
    {
      "epoch": 0.00376,
      "grad_norm": 0.7986123477309184,
      "learning_rate": 0.001128,
      "loss": 5.863,
      "step": 376
    },
    {
      "epoch": 0.00377,
      "grad_norm": 1.034328210427756,
      "learning_rate": 0.001131,
      "loss": 5.8644,
      "step": 377
    },
    {
      "epoch": 0.00378,
      "grad_norm": 1.1394107219722458,
      "learning_rate": 0.001134,
      "loss": 5.8702,
      "step": 378
    },
    {
      "epoch": 0.00379,
      "grad_norm": 0.6778105205682995,
      "learning_rate": 0.001137,
      "loss": 5.8386,
      "step": 379
    },
    {
      "epoch": 0.0038,
      "grad_norm": 0.7735792074169382,
      "learning_rate": 0.00114,
      "loss": 5.8619,
      "step": 380
    },
    {
      "epoch": 0.00381,
      "grad_norm": 0.882590402307869,
      "learning_rate": 0.0011430000000000001,
      "loss": 5.83,
      "step": 381
    },
    {
      "epoch": 0.00382,
      "grad_norm": 0.8860425089624719,
      "learning_rate": 0.001146,
      "loss": 5.8197,
      "step": 382
    },
    {
      "epoch": 0.00383,
      "grad_norm": 1.0396684860073488,
      "learning_rate": 0.001149,
      "loss": 5.836,
      "step": 383
    },
    {
      "epoch": 0.00384,
      "grad_norm": 0.9683520339794217,
      "learning_rate": 0.001152,
      "loss": 5.8386,
      "step": 384
    },
    {
      "epoch": 0.00385,
      "grad_norm": 0.8210443052214895,
      "learning_rate": 0.001155,
      "loss": 5.8215,
      "step": 385
    },
    {
      "epoch": 0.00386,
      "grad_norm": 1.030920073327714,
      "learning_rate": 0.001158,
      "loss": 5.8336,
      "step": 386
    },
    {
      "epoch": 0.00387,
      "grad_norm": 0.8032114385497527,
      "learning_rate": 0.0011610000000000001,
      "loss": 5.8025,
      "step": 387
    },
    {
      "epoch": 0.00388,
      "grad_norm": 0.6803620347459473,
      "learning_rate": 0.001164,
      "loss": 5.7978,
      "step": 388
    },
    {
      "epoch": 0.00389,
      "grad_norm": 0.6752304208768743,
      "learning_rate": 0.001167,
      "loss": 5.8027,
      "step": 389
    },
    {
      "epoch": 0.0039,
      "grad_norm": 0.6054825081153106,
      "learning_rate": 0.00117,
      "loss": 5.7828,
      "step": 390
    },
    {
      "epoch": 0.00391,
      "grad_norm": 0.5156470046541872,
      "learning_rate": 0.001173,
      "loss": 5.7863,
      "step": 391
    },
    {
      "epoch": 0.00392,
      "grad_norm": 0.49609211852516366,
      "learning_rate": 0.001176,
      "loss": 5.7945,
      "step": 392
    },
    {
      "epoch": 0.00393,
      "grad_norm": 0.42817727304572534,
      "learning_rate": 0.0011790000000000001,
      "loss": 5.7664,
      "step": 393
    },
    {
      "epoch": 0.00394,
      "grad_norm": 0.43894767278563757,
      "learning_rate": 0.001182,
      "loss": 5.7539,
      "step": 394
    },
    {
      "epoch": 0.00395,
      "grad_norm": 0.41850660912289844,
      "learning_rate": 0.001185,
      "loss": 5.7366,
      "step": 395
    },
    {
      "epoch": 0.00396,
      "grad_norm": 0.41745423249833347,
      "learning_rate": 0.001188,
      "loss": 5.7516,
      "step": 396
    },
    {
      "epoch": 0.00397,
      "grad_norm": 0.40474070631964676,
      "learning_rate": 0.001191,
      "loss": 5.7433,
      "step": 397
    },
    {
      "epoch": 0.00398,
      "grad_norm": 0.3677722018443306,
      "learning_rate": 0.0011940000000000002,
      "loss": 5.7479,
      "step": 398
    },
    {
      "epoch": 0.00399,
      "grad_norm": 0.3480206547108819,
      "learning_rate": 0.0011970000000000001,
      "loss": 5.7478,
      "step": 399
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.3434828622202681,
      "learning_rate": 0.0012000000000000001,
      "loss": 5.7345,
      "step": 400
    },
    {
      "epoch": 0.00401,
      "grad_norm": 0.34918136204349326,
      "learning_rate": 0.001203,
      "loss": 5.7155,
      "step": 401
    },
    {
      "epoch": 0.00402,
      "grad_norm": 0.30554980038341767,
      "learning_rate": 0.001206,
      "loss": 5.7215,
      "step": 402
    },
    {
      "epoch": 0.00403,
      "grad_norm": 0.38840665384838735,
      "learning_rate": 0.001209,
      "loss": 5.7184,
      "step": 403
    },
    {
      "epoch": 0.00404,
      "grad_norm": 0.5409513056663879,
      "learning_rate": 0.0012120000000000002,
      "loss": 5.714,
      "step": 404
    },
    {
      "epoch": 0.00405,
      "grad_norm": 0.845020924848713,
      "learning_rate": 0.0012150000000000002,
      "loss": 5.7181,
      "step": 405
    },
    {
      "epoch": 0.00406,
      "grad_norm": 1.1911410915070972,
      "learning_rate": 0.0012180000000000001,
      "loss": 5.7205,
      "step": 406
    },
    {
      "epoch": 0.00407,
      "grad_norm": 0.6577283880630926,
      "learning_rate": 0.0012209999999999999,
      "loss": 5.6994,
      "step": 407
    },
    {
      "epoch": 0.00408,
      "grad_norm": 0.7475745975098248,
      "learning_rate": 0.001224,
      "loss": 5.7213,
      "step": 408
    },
    {
      "epoch": 0.00409,
      "grad_norm": 0.9872391126413178,
      "learning_rate": 0.001227,
      "loss": 5.7126,
      "step": 409
    },
    {
      "epoch": 0.0041,
      "grad_norm": 1.1359034721668335,
      "learning_rate": 0.00123,
      "loss": 5.7088,
      "step": 410
    },
    {
      "epoch": 0.00411,
      "grad_norm": 1.3596329145222696,
      "learning_rate": 0.001233,
      "loss": 5.7402,
      "step": 411
    },
    {
      "epoch": 0.00412,
      "grad_norm": 0.7538358474928969,
      "learning_rate": 0.001236,
      "loss": 5.7066,
      "step": 412
    },
    {
      "epoch": 0.00413,
      "grad_norm": 0.9465320539051596,
      "learning_rate": 0.0012389999999999999,
      "loss": 5.7197,
      "step": 413
    },
    {
      "epoch": 0.00414,
      "grad_norm": 0.9262933655624658,
      "learning_rate": 0.001242,
      "loss": 5.6978,
      "step": 414
    },
    {
      "epoch": 0.00415,
      "grad_norm": 1.1564175286146172,
      "learning_rate": 0.001245,
      "loss": 5.7105,
      "step": 415
    },
    {
      "epoch": 0.00416,
      "grad_norm": 1.1001247072345506,
      "learning_rate": 0.001248,
      "loss": 5.6929,
      "step": 416
    },
    {
      "epoch": 0.00417,
      "grad_norm": 1.0416153435685582,
      "learning_rate": 0.001251,
      "loss": 5.7199,
      "step": 417
    },
    {
      "epoch": 0.00418,
      "grad_norm": 1.0281555694116995,
      "learning_rate": 0.001254,
      "loss": 5.6999,
      "step": 418
    },
    {
      "epoch": 0.00419,
      "grad_norm": 1.1154617103247704,
      "learning_rate": 0.0012569999999999999,
      "loss": 5.6876,
      "step": 419
    },
    {
      "epoch": 0.0042,
      "grad_norm": 0.9999912825556322,
      "learning_rate": 0.00126,
      "loss": 5.7045,
      "step": 420
    },
    {
      "epoch": 0.00421,
      "grad_norm": 0.9729638313238949,
      "learning_rate": 0.001263,
      "loss": 5.6933,
      "step": 421
    },
    {
      "epoch": 0.00422,
      "grad_norm": 1.0297208241186608,
      "learning_rate": 0.001266,
      "loss": 5.6894,
      "step": 422
    },
    {
      "epoch": 0.00423,
      "grad_norm": 0.8381062456476874,
      "learning_rate": 0.001269,
      "loss": 5.6811,
      "step": 423
    },
    {
      "epoch": 0.00424,
      "grad_norm": 0.73825928337582,
      "learning_rate": 0.001272,
      "loss": 5.6773,
      "step": 424
    },
    {
      "epoch": 0.00425,
      "grad_norm": 0.8151693610653118,
      "learning_rate": 0.001275,
      "loss": 5.6815,
      "step": 425
    },
    {
      "epoch": 0.00426,
      "grad_norm": 0.9877723869544494,
      "learning_rate": 0.001278,
      "loss": 5.6881,
      "step": 426
    },
    {
      "epoch": 0.00427,
      "grad_norm": 0.8875063218226354,
      "learning_rate": 0.001281,
      "loss": 5.6676,
      "step": 427
    },
    {
      "epoch": 0.00428,
      "grad_norm": 0.8761493100474917,
      "learning_rate": 0.001284,
      "loss": 5.6437,
      "step": 428
    },
    {
      "epoch": 0.00429,
      "grad_norm": 1.109439755296158,
      "learning_rate": 0.001287,
      "loss": 5.668,
      "step": 429
    },
    {
      "epoch": 0.0043,
      "grad_norm": 0.8726361797071612,
      "learning_rate": 0.00129,
      "loss": 5.6611,
      "step": 430
    },
    {
      "epoch": 0.00431,
      "grad_norm": 0.5601358924900582,
      "learning_rate": 0.001293,
      "loss": 5.6342,
      "step": 431
    },
    {
      "epoch": 0.00432,
      "grad_norm": 0.5940988397687699,
      "learning_rate": 0.001296,
      "loss": 5.6448,
      "step": 432
    },
    {
      "epoch": 0.00433,
      "grad_norm": 0.480538337754196,
      "learning_rate": 0.001299,
      "loss": 5.6336,
      "step": 433
    },
    {
      "epoch": 0.00434,
      "grad_norm": 0.5069699406038057,
      "learning_rate": 0.001302,
      "loss": 5.6331,
      "step": 434
    },
    {
      "epoch": 0.00435,
      "grad_norm": 0.417220569666801,
      "learning_rate": 0.001305,
      "loss": 5.6227,
      "step": 435
    },
    {
      "epoch": 0.00436,
      "grad_norm": 0.46597565539240443,
      "learning_rate": 0.001308,
      "loss": 5.6039,
      "step": 436
    },
    {
      "epoch": 0.00437,
      "grad_norm": 0.37605303006482044,
      "learning_rate": 0.001311,
      "loss": 5.6161,
      "step": 437
    },
    {
      "epoch": 0.00438,
      "grad_norm": 0.3526339213940271,
      "learning_rate": 0.001314,
      "loss": 5.5977,
      "step": 438
    },
    {
      "epoch": 0.00439,
      "grad_norm": 0.3618369277094543,
      "learning_rate": 0.001317,
      "loss": 5.6246,
      "step": 439
    },
    {
      "epoch": 0.0044,
      "grad_norm": 0.3223158135938896,
      "learning_rate": 0.00132,
      "loss": 5.5938,
      "step": 440
    },
    {
      "epoch": 0.00441,
      "grad_norm": 0.3386640445759432,
      "learning_rate": 0.001323,
      "loss": 5.5905,
      "step": 441
    },
    {
      "epoch": 0.00442,
      "grad_norm": 0.3397360216396013,
      "learning_rate": 0.0013260000000000001,
      "loss": 5.5838,
      "step": 442
    },
    {
      "epoch": 0.00443,
      "grad_norm": 0.34532354892574607,
      "learning_rate": 0.001329,
      "loss": 5.5832,
      "step": 443
    },
    {
      "epoch": 0.00444,
      "grad_norm": 0.37928556611065656,
      "learning_rate": 0.001332,
      "loss": 5.5853,
      "step": 444
    },
    {
      "epoch": 0.00445,
      "grad_norm": 0.44947894711961484,
      "learning_rate": 0.001335,
      "loss": 5.5723,
      "step": 445
    },
    {
      "epoch": 0.00446,
      "grad_norm": 0.5239308309237933,
      "learning_rate": 0.001338,
      "loss": 5.5751,
      "step": 446
    },
    {
      "epoch": 0.00447,
      "grad_norm": 0.60889528761374,
      "learning_rate": 0.001341,
      "loss": 5.5777,
      "step": 447
    },
    {
      "epoch": 0.00448,
      "grad_norm": 0.6150556040535831,
      "learning_rate": 0.0013440000000000001,
      "loss": 5.561,
      "step": 448
    },
    {
      "epoch": 0.00449,
      "grad_norm": 0.5444850551876294,
      "learning_rate": 0.001347,
      "loss": 5.5623,
      "step": 449
    },
    {
      "epoch": 0.0045,
      "grad_norm": 0.6163688303714219,
      "learning_rate": 0.00135,
      "loss": 5.5617,
      "step": 450
    },
    {
      "epoch": 0.00451,
      "grad_norm": 0.7972728836658292,
      "learning_rate": 0.001353,
      "loss": 5.5614,
      "step": 451
    },
    {
      "epoch": 0.00452,
      "grad_norm": 0.7711764764543457,
      "learning_rate": 0.001356,
      "loss": 5.5454,
      "step": 452
    },
    {
      "epoch": 0.00453,
      "grad_norm": 0.7702356986189732,
      "learning_rate": 0.001359,
      "loss": 5.5379,
      "step": 453
    },
    {
      "epoch": 0.00454,
      "grad_norm": 1.0838160417982272,
      "learning_rate": 0.0013620000000000001,
      "loss": 5.5686,
      "step": 454
    },
    {
      "epoch": 0.00455,
      "grad_norm": 1.152655585803101,
      "learning_rate": 0.0013650000000000001,
      "loss": 5.5757,
      "step": 455
    },
    {
      "epoch": 0.00456,
      "grad_norm": 1.2776355160689266,
      "learning_rate": 0.001368,
      "loss": 5.5831,
      "step": 456
    },
    {
      "epoch": 0.00457,
      "grad_norm": 1.091292995937963,
      "learning_rate": 0.001371,
      "loss": 5.5727,
      "step": 457
    },
    {
      "epoch": 0.00458,
      "grad_norm": 0.8963405103823251,
      "learning_rate": 0.001374,
      "loss": 5.5735,
      "step": 458
    },
    {
      "epoch": 0.00459,
      "grad_norm": 1.0168648046101516,
      "learning_rate": 0.0013770000000000002,
      "loss": 5.5669,
      "step": 459
    },
    {
      "epoch": 0.0046,
      "grad_norm": 1.296131601782866,
      "learning_rate": 0.0013800000000000002,
      "loss": 5.5665,
      "step": 460
    },
    {
      "epoch": 0.00461,
      "grad_norm": 0.9634420565591739,
      "learning_rate": 0.0013830000000000001,
      "loss": 5.556,
      "step": 461
    },
    {
      "epoch": 0.00462,
      "grad_norm": 0.9383281224355017,
      "learning_rate": 0.001386,
      "loss": 5.5634,
      "step": 462
    },
    {
      "epoch": 0.00463,
      "grad_norm": 1.2569585597421309,
      "learning_rate": 0.001389,
      "loss": 5.5542,
      "step": 463
    },
    {
      "epoch": 0.00464,
      "grad_norm": 0.9874595595654581,
      "learning_rate": 0.001392,
      "loss": 5.5689,
      "step": 464
    },
    {
      "epoch": 0.00465,
      "grad_norm": 1.085595749506429,
      "learning_rate": 0.0013950000000000002,
      "loss": 5.5385,
      "step": 465
    },
    {
      "epoch": 0.00466,
      "grad_norm": 1.0673943770446899,
      "learning_rate": 0.0013980000000000002,
      "loss": 5.5603,
      "step": 466
    },
    {
      "epoch": 0.00467,
      "grad_norm": 0.8139501043376736,
      "learning_rate": 0.0014010000000000001,
      "loss": 5.5432,
      "step": 467
    },
    {
      "epoch": 0.00468,
      "grad_norm": 0.7494382793960519,
      "learning_rate": 0.001404,
      "loss": 5.5245,
      "step": 468
    },
    {
      "epoch": 0.00469,
      "grad_norm": 0.7634992086588068,
      "learning_rate": 0.001407,
      "loss": 5.5282,
      "step": 469
    },
    {
      "epoch": 0.0047,
      "grad_norm": 0.8018093758476836,
      "learning_rate": 0.00141,
      "loss": 5.5404,
      "step": 470
    },
    {
      "epoch": 0.00471,
      "grad_norm": 0.7418690809708749,
      "learning_rate": 0.001413,
      "loss": 5.5115,
      "step": 471
    },
    {
      "epoch": 0.00472,
      "grad_norm": 0.7355325431039438,
      "learning_rate": 0.001416,
      "loss": 5.5216,
      "step": 472
    },
    {
      "epoch": 0.00473,
      "grad_norm": 0.709026539269664,
      "learning_rate": 0.001419,
      "loss": 5.5305,
      "step": 473
    },
    {
      "epoch": 0.00474,
      "grad_norm": 0.5742329758009745,
      "learning_rate": 0.0014219999999999999,
      "loss": 5.5064,
      "step": 474
    },
    {
      "epoch": 0.00475,
      "grad_norm": 0.5859758403725885,
      "learning_rate": 0.001425,
      "loss": 5.4971,
      "step": 475
    },
    {
      "epoch": 0.00476,
      "grad_norm": 0.6365922795308678,
      "learning_rate": 0.001428,
      "loss": 5.5308,
      "step": 476
    },
    {
      "epoch": 0.00477,
      "grad_norm": 0.6539516343537074,
      "learning_rate": 0.001431,
      "loss": 5.4846,
      "step": 477
    },
    {
      "epoch": 0.00478,
      "grad_norm": 0.6446859909585969,
      "learning_rate": 0.001434,
      "loss": 5.5074,
      "step": 478
    },
    {
      "epoch": 0.00479,
      "grad_norm": 0.7791938725908187,
      "learning_rate": 0.001437,
      "loss": 5.4897,
      "step": 479
    },
    {
      "epoch": 0.0048,
      "grad_norm": 0.7608695568360718,
      "learning_rate": 0.0014399999999999999,
      "loss": 5.4855,
      "step": 480
    },
    {
      "epoch": 0.00481,
      "grad_norm": 0.5435552532069989,
      "learning_rate": 0.001443,
      "loss": 5.4813,
      "step": 481
    },
    {
      "epoch": 0.00482,
      "grad_norm": 0.4961185149512517,
      "learning_rate": 0.001446,
      "loss": 5.4538,
      "step": 482
    },
    {
      "epoch": 0.00483,
      "grad_norm": 0.5120902463904886,
      "learning_rate": 0.001449,
      "loss": 5.4636,
      "step": 483
    },
    {
      "epoch": 0.00484,
      "grad_norm": 0.418197369903841,
      "learning_rate": 0.001452,
      "loss": 5.464,
      "step": 484
    },
    {
      "epoch": 0.00485,
      "grad_norm": 0.36311406822078424,
      "learning_rate": 0.001455,
      "loss": 5.4671,
      "step": 485
    },
    {
      "epoch": 0.00486,
      "grad_norm": 0.3913777576995821,
      "learning_rate": 0.001458,
      "loss": 5.4393,
      "step": 486
    },
    {
      "epoch": 0.00487,
      "grad_norm": 0.36874474197662527,
      "learning_rate": 0.001461,
      "loss": 5.4517,
      "step": 487
    },
    {
      "epoch": 0.00488,
      "grad_norm": 0.41593519822402414,
      "learning_rate": 0.001464,
      "loss": 5.43,
      "step": 488
    },
    {
      "epoch": 0.00489,
      "grad_norm": 0.4051383375955623,
      "learning_rate": 0.001467,
      "loss": 5.4204,
      "step": 489
    },
    {
      "epoch": 0.0049,
      "grad_norm": 0.4948319932753325,
      "learning_rate": 0.00147,
      "loss": 5.434,
      "step": 490
    },
    {
      "epoch": 0.00491,
      "grad_norm": 0.6887805434617323,
      "learning_rate": 0.001473,
      "loss": 5.4357,
      "step": 491
    },
    {
      "epoch": 0.00492,
      "grad_norm": 0.912047432270828,
      "learning_rate": 0.001476,
      "loss": 5.4432,
      "step": 492
    },
    {
      "epoch": 0.00493,
      "grad_norm": 0.8950009277905591,
      "learning_rate": 0.001479,
      "loss": 5.4415,
      "step": 493
    },
    {
      "epoch": 0.00494,
      "grad_norm": 0.807683799932231,
      "learning_rate": 0.001482,
      "loss": 5.4427,
      "step": 494
    },
    {
      "epoch": 0.00495,
      "grad_norm": 0.976722169908224,
      "learning_rate": 0.001485,
      "loss": 5.4601,
      "step": 495
    },
    {
      "epoch": 0.00496,
      "grad_norm": 0.7224780529872387,
      "learning_rate": 0.001488,
      "loss": 5.4314,
      "step": 496
    },
    {
      "epoch": 0.00497,
      "grad_norm": 0.7139613919522917,
      "learning_rate": 0.001491,
      "loss": 5.4172,
      "step": 497
    },
    {
      "epoch": 0.00498,
      "grad_norm": 0.7556637282468179,
      "learning_rate": 0.001494,
      "loss": 5.4443,
      "step": 498
    },
    {
      "epoch": 0.00499,
      "grad_norm": 0.8519321120342865,
      "learning_rate": 0.001497,
      "loss": 5.4223,
      "step": 499
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.9868888032280079,
      "learning_rate": 0.0015,
      "loss": 5.4308,
      "step": 500
    },
    {
      "epoch": 0.00501,
      "grad_norm": 1.2028724714017198,
      "learning_rate": 0.001503,
      "loss": 5.4458,
      "step": 501
    },
    {
      "epoch": 0.00502,
      "grad_norm": 0.9548534640519003,
      "learning_rate": 0.001506,
      "loss": 5.4405,
      "step": 502
    },
    {
      "epoch": 0.00503,
      "grad_norm": 0.9423651398338494,
      "learning_rate": 0.0015090000000000001,
      "loss": 5.4484,
      "step": 503
    },
    {
      "epoch": 0.00504,
      "grad_norm": 0.9672711526274779,
      "learning_rate": 0.001512,
      "loss": 5.4328,
      "step": 504
    },
    {
      "epoch": 0.00505,
      "grad_norm": 0.9474889600476256,
      "learning_rate": 0.001515,
      "loss": 5.4283,
      "step": 505
    },
    {
      "epoch": 0.00506,
      "grad_norm": 1.3039451405080307,
      "learning_rate": 0.001518,
      "loss": 5.4432,
      "step": 506
    },
    {
      "epoch": 0.00507,
      "grad_norm": 1.2269782118632737,
      "learning_rate": 0.001521,
      "loss": 5.4452,
      "step": 507
    },
    {
      "epoch": 0.00508,
      "grad_norm": 0.8597426683067237,
      "learning_rate": 0.001524,
      "loss": 5.4268,
      "step": 508
    },
    {
      "epoch": 0.00509,
      "grad_norm": 0.9388657090373522,
      "learning_rate": 0.0015270000000000001,
      "loss": 5.4178,
      "step": 509
    },
    {
      "epoch": 0.0051,
      "grad_norm": 0.9594427813189665,
      "learning_rate": 0.0015300000000000001,
      "loss": 5.4356,
      "step": 510
    },
    {
      "epoch": 0.00511,
      "grad_norm": 1.04563577032056,
      "learning_rate": 0.001533,
      "loss": 5.4212,
      "step": 511
    },
    {
      "epoch": 0.00512,
      "grad_norm": 0.733703407645156,
      "learning_rate": 0.001536,
      "loss": 5.4003,
      "step": 512
    },
    {
      "epoch": 0.00513,
      "grad_norm": 0.8415210942026606,
      "learning_rate": 0.001539,
      "loss": 5.423,
      "step": 513
    },
    {
      "epoch": 0.00514,
      "grad_norm": 0.8791751992621939,
      "learning_rate": 0.001542,
      "loss": 5.4064,
      "step": 514
    },
    {
      "epoch": 0.00515,
      "grad_norm": 0.8161499995578689,
      "learning_rate": 0.0015450000000000001,
      "loss": 5.4094,
      "step": 515
    },
    {
      "epoch": 0.00516,
      "grad_norm": 0.878383695319614,
      "learning_rate": 0.0015480000000000001,
      "loss": 5.4087,
      "step": 516
    },
    {
      "epoch": 0.00517,
      "grad_norm": 0.9768725869756134,
      "learning_rate": 0.001551,
      "loss": 5.4055,
      "step": 517
    },
    {
      "epoch": 0.00518,
      "grad_norm": 0.8865017723772849,
      "learning_rate": 0.001554,
      "loss": 5.3907,
      "step": 518
    },
    {
      "epoch": 0.00519,
      "grad_norm": 0.8308797688973832,
      "learning_rate": 0.001557,
      "loss": 5.3905,
      "step": 519
    },
    {
      "epoch": 0.0052,
      "grad_norm": 0.6978413162257922,
      "learning_rate": 0.0015600000000000002,
      "loss": 5.3938,
      "step": 520
    },
    {
      "epoch": 0.00521,
      "grad_norm": 0.6562689530690187,
      "learning_rate": 0.0015630000000000002,
      "loss": 5.3676,
      "step": 521
    },
    {
      "epoch": 0.00522,
      "grad_norm": 0.5577523148431155,
      "learning_rate": 0.0015660000000000001,
      "loss": 5.3673,
      "step": 522
    },
    {
      "epoch": 0.00523,
      "grad_norm": 0.5298728018270966,
      "learning_rate": 0.001569,
      "loss": 5.3784,
      "step": 523
    },
    {
      "epoch": 0.00524,
      "grad_norm": 0.44216623389663734,
      "learning_rate": 0.001572,
      "loss": 5.3811,
      "step": 524
    },
    {
      "epoch": 0.00525,
      "grad_norm": 0.3702182111689363,
      "learning_rate": 0.001575,
      "loss": 5.369,
      "step": 525
    },
    {
      "epoch": 0.00526,
      "grad_norm": 0.40289552198632295,
      "learning_rate": 0.0015780000000000002,
      "loss": 5.3444,
      "step": 526
    },
    {
      "epoch": 0.00527,
      "grad_norm": 0.38284579814689895,
      "learning_rate": 0.0015810000000000002,
      "loss": 5.3551,
      "step": 527
    },
    {
      "epoch": 0.00528,
      "grad_norm": 0.33950473031510653,
      "learning_rate": 0.0015840000000000001,
      "loss": 5.3339,
      "step": 528
    },
    {
      "epoch": 0.00529,
      "grad_norm": 0.3777758983585419,
      "learning_rate": 0.001587,
      "loss": 5.343,
      "step": 529
    },
    {
      "epoch": 0.0053,
      "grad_norm": 0.4257451161382566,
      "learning_rate": 0.00159,
      "loss": 5.3483,
      "step": 530
    },
    {
      "epoch": 0.00531,
      "grad_norm": 0.5472045550610978,
      "learning_rate": 0.001593,
      "loss": 5.3387,
      "step": 531
    },
    {
      "epoch": 0.00532,
      "grad_norm": 0.696856419507981,
      "learning_rate": 0.0015960000000000002,
      "loss": 5.3455,
      "step": 532
    },
    {
      "epoch": 0.00533,
      "grad_norm": 0.7975941430607876,
      "learning_rate": 0.0015990000000000002,
      "loss": 5.3447,
      "step": 533
    },
    {
      "epoch": 0.00534,
      "grad_norm": 0.6576622452461942,
      "learning_rate": 0.0016020000000000001,
      "loss": 5.3339,
      "step": 534
    },
    {
      "epoch": 0.00535,
      "grad_norm": 0.5879483820814444,
      "learning_rate": 0.001605,
      "loss": 5.3347,
      "step": 535
    },
    {
      "epoch": 0.00536,
      "grad_norm": 0.8212636469840171,
      "learning_rate": 0.001608,
      "loss": 5.3454,
      "step": 536
    },
    {
      "epoch": 0.00537,
      "grad_norm": 0.7740232031924225,
      "learning_rate": 0.0016110000000000002,
      "loss": 5.3418,
      "step": 537
    },
    {
      "epoch": 0.00538,
      "grad_norm": 0.683098129060214,
      "learning_rate": 0.0016140000000000002,
      "loss": 5.3143,
      "step": 538
    },
    {
      "epoch": 0.00539,
      "grad_norm": 0.9170551975741953,
      "learning_rate": 0.0016170000000000002,
      "loss": 5.3245,
      "step": 539
    },
    {
      "epoch": 0.0054,
      "grad_norm": 0.765080131514484,
      "learning_rate": 0.0016200000000000001,
      "loss": 5.3262,
      "step": 540
    },
    {
      "epoch": 0.00541,
      "grad_norm": 0.658457026305436,
      "learning_rate": 0.001623,
      "loss": 5.3137,
      "step": 541
    },
    {
      "epoch": 0.00542,
      "grad_norm": 0.601942869875084,
      "learning_rate": 0.001626,
      "loss": 5.3315,
      "step": 542
    },
    {
      "epoch": 0.00543,
      "grad_norm": 0.6751097730454854,
      "learning_rate": 0.0016290000000000002,
      "loss": 5.2998,
      "step": 543
    },
    {
      "epoch": 0.00544,
      "grad_norm": 0.6943293389301006,
      "learning_rate": 0.0016320000000000002,
      "loss": 5.3191,
      "step": 544
    },
    {
      "epoch": 0.00545,
      "grad_norm": 0.757591621302123,
      "learning_rate": 0.0016350000000000002,
      "loss": 5.3293,
      "step": 545
    },
    {
      "epoch": 0.00546,
      "grad_norm": 0.825685624372282,
      "learning_rate": 0.0016380000000000001,
      "loss": 5.3061,
      "step": 546
    },
    {
      "epoch": 0.00547,
      "grad_norm": 0.8411824339962438,
      "learning_rate": 0.001641,
      "loss": 5.3051,
      "step": 547
    },
    {
      "epoch": 0.00548,
      "grad_norm": 0.8225125547688507,
      "learning_rate": 0.001644,
      "loss": 5.303,
      "step": 548
    },
    {
      "epoch": 0.00549,
      "grad_norm": 0.7692066087665821,
      "learning_rate": 0.0016470000000000002,
      "loss": 5.3016,
      "step": 549
    },
    {
      "epoch": 0.0055,
      "grad_norm": 0.7541680263658305,
      "learning_rate": 0.0016500000000000002,
      "loss": 5.3099,
      "step": 550
    },
    {
      "epoch": 0.00551,
      "grad_norm": 0.8632990593818363,
      "learning_rate": 0.0016530000000000002,
      "loss": 5.3123,
      "step": 551
    },
    {
      "epoch": 0.00552,
      "grad_norm": 0.9083583396116485,
      "learning_rate": 0.0016560000000000001,
      "loss": 5.3063,
      "step": 552
    },
    {
      "epoch": 0.00553,
      "grad_norm": 0.9125681148017897,
      "learning_rate": 0.001659,
      "loss": 5.304,
      "step": 553
    },
    {
      "epoch": 0.00554,
      "grad_norm": 0.9360850923631763,
      "learning_rate": 0.0016620000000000003,
      "loss": 5.3021,
      "step": 554
    },
    {
      "epoch": 0.00555,
      "grad_norm": 0.9519012966124738,
      "learning_rate": 0.0016650000000000002,
      "loss": 5.3154,
      "step": 555
    },
    {
      "epoch": 0.00556,
      "grad_norm": 0.9297335237671114,
      "learning_rate": 0.0016680000000000002,
      "loss": 5.2962,
      "step": 556
    },
    {
      "epoch": 0.00557,
      "grad_norm": 1.1187109821510643,
      "learning_rate": 0.0016710000000000002,
      "loss": 5.2935,
      "step": 557
    },
    {
      "epoch": 0.00558,
      "grad_norm": 0.9868599839032521,
      "learning_rate": 0.0016740000000000001,
      "loss": 5.3219,
      "step": 558
    },
    {
      "epoch": 0.00559,
      "grad_norm": 0.8786610310513777,
      "learning_rate": 0.001677,
      "loss": 5.2938,
      "step": 559
    },
    {
      "epoch": 0.0056,
      "grad_norm": 0.9118163270219521,
      "learning_rate": 0.0016800000000000003,
      "loss": 5.2767,
      "step": 560
    },
    {
      "epoch": 0.00561,
      "grad_norm": 0.8649386991384733,
      "learning_rate": 0.0016830000000000003,
      "loss": 5.3004,
      "step": 561
    },
    {
      "epoch": 0.00562,
      "grad_norm": 0.5982058914442256,
      "learning_rate": 0.0016860000000000002,
      "loss": 5.279,
      "step": 562
    },
    {
      "epoch": 0.00563,
      "grad_norm": 0.5830709316445766,
      "learning_rate": 0.001689,
      "loss": 5.2821,
      "step": 563
    },
    {
      "epoch": 0.00564,
      "grad_norm": 0.5564672250442253,
      "learning_rate": 0.001692,
      "loss": 5.2582,
      "step": 564
    },
    {
      "epoch": 0.00565,
      "grad_norm": 0.5903966881939692,
      "learning_rate": 0.001695,
      "loss": 5.2619,
      "step": 565
    },
    {
      "epoch": 0.00566,
      "grad_norm": 0.5777761648359326,
      "learning_rate": 0.0016979999999999999,
      "loss": 5.2552,
      "step": 566
    },
    {
      "epoch": 0.00567,
      "grad_norm": 0.6069030603134064,
      "learning_rate": 0.0017009999999999998,
      "loss": 5.2491,
      "step": 567
    },
    {
      "epoch": 0.00568,
      "grad_norm": 0.5693132087018719,
      "learning_rate": 0.0017039999999999998,
      "loss": 5.2604,
      "step": 568
    },
    {
      "epoch": 0.00569,
      "grad_norm": 0.47662406778838745,
      "learning_rate": 0.001707,
      "loss": 5.2359,
      "step": 569
    },
    {
      "epoch": 0.0057,
      "grad_norm": 0.49231021705037487,
      "learning_rate": 0.00171,
      "loss": 5.253,
      "step": 570
    },
    {
      "epoch": 0.00571,
      "grad_norm": 0.4167352661720621,
      "learning_rate": 0.001713,
      "loss": 5.2491,
      "step": 571
    },
    {
      "epoch": 0.00572,
      "grad_norm": 0.39011854138215074,
      "learning_rate": 0.0017159999999999999,
      "loss": 5.2258,
      "step": 572
    },
    {
      "epoch": 0.00573,
      "grad_norm": 0.45971389455263184,
      "learning_rate": 0.0017189999999999998,
      "loss": 5.2313,
      "step": 573
    },
    {
      "epoch": 0.00574,
      "grad_norm": 0.4643694238461635,
      "learning_rate": 0.001722,
      "loss": 5.2399,
      "step": 574
    },
    {
      "epoch": 0.00575,
      "grad_norm": 0.4439840100405838,
      "learning_rate": 0.001725,
      "loss": 5.2168,
      "step": 575
    },
    {
      "epoch": 0.00576,
      "grad_norm": 0.4794795740566699,
      "learning_rate": 0.001728,
      "loss": 5.2109,
      "step": 576
    },
    {
      "epoch": 0.00577,
      "grad_norm": 0.5445912068882687,
      "learning_rate": 0.001731,
      "loss": 5.2354,
      "step": 577
    },
    {
      "epoch": 0.00578,
      "grad_norm": 0.5881619532039025,
      "learning_rate": 0.0017339999999999999,
      "loss": 5.2156,
      "step": 578
    },
    {
      "epoch": 0.00579,
      "grad_norm": 0.6688964155004951,
      "learning_rate": 0.0017369999999999998,
      "loss": 5.2064,
      "step": 579
    },
    {
      "epoch": 0.0058,
      "grad_norm": 0.6446644513052245,
      "learning_rate": 0.00174,
      "loss": 5.2223,
      "step": 580
    },
    {
      "epoch": 0.00581,
      "grad_norm": 0.7385299343409043,
      "learning_rate": 0.001743,
      "loss": 5.2222,
      "step": 581
    },
    {
      "epoch": 0.00582,
      "grad_norm": 1.0486135149054512,
      "learning_rate": 0.001746,
      "loss": 5.2114,
      "step": 582
    },
    {
      "epoch": 0.00583,
      "grad_norm": 0.9184260585056472,
      "learning_rate": 0.001749,
      "loss": 5.2374,
      "step": 583
    },
    {
      "epoch": 0.00584,
      "grad_norm": 0.7724931135788974,
      "learning_rate": 0.0017519999999999999,
      "loss": 5.2235,
      "step": 584
    },
    {
      "epoch": 0.00585,
      "grad_norm": 0.903942587317279,
      "learning_rate": 0.0017549999999999998,
      "loss": 5.2218,
      "step": 585
    },
    {
      "epoch": 0.00586,
      "grad_norm": 0.8496888678331875,
      "learning_rate": 0.001758,
      "loss": 5.2272,
      "step": 586
    },
    {
      "epoch": 0.00587,
      "grad_norm": 0.8580070219006531,
      "learning_rate": 0.001761,
      "loss": 5.2094,
      "step": 587
    },
    {
      "epoch": 0.00588,
      "grad_norm": 1.0100984711915582,
      "learning_rate": 0.001764,
      "loss": 5.233,
      "step": 588
    },
    {
      "epoch": 0.00589,
      "grad_norm": 0.9311738464832717,
      "learning_rate": 0.001767,
      "loss": 5.2289,
      "step": 589
    },
    {
      "epoch": 0.0059,
      "grad_norm": 0.9363758859588256,
      "learning_rate": 0.0017699999999999999,
      "loss": 5.2155,
      "step": 590
    },
    {
      "epoch": 0.00591,
      "grad_norm": 0.9695025163620248,
      "learning_rate": 0.001773,
      "loss": 5.2165,
      "step": 591
    },
    {
      "epoch": 0.00592,
      "grad_norm": 0.9805216711841677,
      "learning_rate": 0.001776,
      "loss": 5.2417,
      "step": 592
    },
    {
      "epoch": 0.00593,
      "grad_norm": 0.9827241162755458,
      "learning_rate": 0.001779,
      "loss": 5.2392,
      "step": 593
    },
    {
      "epoch": 0.00594,
      "grad_norm": 1.1445433178010636,
      "learning_rate": 0.001782,
      "loss": 5.2522,
      "step": 594
    },
    {
      "epoch": 0.00595,
      "grad_norm": 0.950699335886284,
      "learning_rate": 0.001785,
      "loss": 5.2351,
      "step": 595
    },
    {
      "epoch": 0.00596,
      "grad_norm": 0.8373647963925929,
      "learning_rate": 0.0017879999999999999,
      "loss": 5.2269,
      "step": 596
    },
    {
      "epoch": 0.00597,
      "grad_norm": 0.9799263768822638,
      "learning_rate": 0.001791,
      "loss": 5.2118,
      "step": 597
    },
    {
      "epoch": 0.00598,
      "grad_norm": 0.9413379648187816,
      "learning_rate": 0.001794,
      "loss": 5.2215,
      "step": 598
    },
    {
      "epoch": 0.00599,
      "grad_norm": 0.8392983563516706,
      "learning_rate": 0.001797,
      "loss": 5.2191,
      "step": 599
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.8414286288934597,
      "learning_rate": 0.0018,
      "loss": 5.2134,
      "step": 600
    },
    {
      "epoch": 0.00601,
      "grad_norm": 0.8993201270126511,
      "learning_rate": 0.001803,
      "loss": 5.2089,
      "step": 601
    },
    {
      "epoch": 0.00602,
      "grad_norm": 0.8678008919316418,
      "learning_rate": 0.0018059999999999999,
      "loss": 5.1957,
      "step": 602
    },
    {
      "epoch": 0.00603,
      "grad_norm": 0.6967382011990112,
      "learning_rate": 0.001809,
      "loss": 5.2122,
      "step": 603
    },
    {
      "epoch": 0.00604,
      "grad_norm": 0.5312401482691863,
      "learning_rate": 0.001812,
      "loss": 5.1788,
      "step": 604
    },
    {
      "epoch": 0.00605,
      "grad_norm": 0.5795052729398412,
      "learning_rate": 0.001815,
      "loss": 5.1915,
      "step": 605
    },
    {
      "epoch": 0.00606,
      "grad_norm": 0.6345039250131549,
      "learning_rate": 0.001818,
      "loss": 5.1894,
      "step": 606
    },
    {
      "epoch": 0.00607,
      "grad_norm": 0.6716049737606092,
      "learning_rate": 0.001821,
      "loss": 5.1916,
      "step": 607
    },
    {
      "epoch": 0.00608,
      "grad_norm": 0.631863512626499,
      "learning_rate": 0.001824,
      "loss": 5.1762,
      "step": 608
    },
    {
      "epoch": 0.00609,
      "grad_norm": 0.5345713642167306,
      "learning_rate": 0.001827,
      "loss": 5.177,
      "step": 609
    },
    {
      "epoch": 0.0061,
      "grad_norm": 0.455626202663485,
      "learning_rate": 0.00183,
      "loss": 5.1434,
      "step": 610
    },
    {
      "epoch": 0.00611,
      "grad_norm": 0.46197362206927406,
      "learning_rate": 0.001833,
      "loss": 5.1489,
      "step": 611
    },
    {
      "epoch": 0.00612,
      "grad_norm": 0.3943653846553885,
      "learning_rate": 0.001836,
      "loss": 5.1488,
      "step": 612
    },
    {
      "epoch": 0.00613,
      "grad_norm": 0.4053887102586243,
      "learning_rate": 0.001839,
      "loss": 5.1409,
      "step": 613
    },
    {
      "epoch": 0.00614,
      "grad_norm": 0.44363743610311057,
      "learning_rate": 0.001842,
      "loss": 5.1611,
      "step": 614
    },
    {
      "epoch": 0.00615,
      "grad_norm": 0.40025757691173514,
      "learning_rate": 0.001845,
      "loss": 5.145,
      "step": 615
    },
    {
      "epoch": 0.00616,
      "grad_norm": 0.4067321452673224,
      "learning_rate": 0.001848,
      "loss": 5.1299,
      "step": 616
    },
    {
      "epoch": 0.00617,
      "grad_norm": 0.42403873536126996,
      "learning_rate": 0.001851,
      "loss": 5.1397,
      "step": 617
    },
    {
      "epoch": 0.00618,
      "grad_norm": 0.44192084147381183,
      "learning_rate": 0.001854,
      "loss": 5.1265,
      "step": 618
    },
    {
      "epoch": 0.00619,
      "grad_norm": 0.5104655865853382,
      "learning_rate": 0.001857,
      "loss": 5.1299,
      "step": 619
    },
    {
      "epoch": 0.0062,
      "grad_norm": 0.6226145292933671,
      "learning_rate": 0.00186,
      "loss": 5.118,
      "step": 620
    },
    {
      "epoch": 0.00621,
      "grad_norm": 0.745237667035521,
      "learning_rate": 0.001863,
      "loss": 5.1237,
      "step": 621
    },
    {
      "epoch": 0.00622,
      "grad_norm": 0.7986865810762169,
      "learning_rate": 0.001866,
      "loss": 5.1267,
      "step": 622
    },
    {
      "epoch": 0.00623,
      "grad_norm": 0.7723208155309241,
      "learning_rate": 0.001869,
      "loss": 5.1084,
      "step": 623
    },
    {
      "epoch": 0.00624,
      "grad_norm": 0.8171878766946273,
      "learning_rate": 0.001872,
      "loss": 5.1246,
      "step": 624
    },
    {
      "epoch": 0.00625,
      "grad_norm": 0.7607424856276187,
      "learning_rate": 0.001875,
      "loss": 5.1359,
      "step": 625
    },
    {
      "epoch": 0.00626,
      "grad_norm": 0.7005729458198662,
      "learning_rate": 0.0018780000000000001,
      "loss": 5.1104,
      "step": 626
    },
    {
      "epoch": 0.00627,
      "grad_norm": 0.6735260479679158,
      "learning_rate": 0.001881,
      "loss": 5.1015,
      "step": 627
    },
    {
      "epoch": 0.00628,
      "grad_norm": 0.780486998163919,
      "learning_rate": 0.001884,
      "loss": 5.1349,
      "step": 628
    },
    {
      "epoch": 0.00629,
      "grad_norm": 0.8746830311438225,
      "learning_rate": 0.001887,
      "loss": 5.1097,
      "step": 629
    },
    {
      "epoch": 0.0063,
      "grad_norm": 0.9536711950620466,
      "learning_rate": 0.00189,
      "loss": 5.1137,
      "step": 630
    },
    {
      "epoch": 0.00631,
      "grad_norm": 0.8628106812141149,
      "learning_rate": 0.0018930000000000002,
      "loss": 5.1035,
      "step": 631
    },
    {
      "epoch": 0.00632,
      "grad_norm": 0.9320890333045916,
      "learning_rate": 0.0018960000000000001,
      "loss": 5.1281,
      "step": 632
    },
    {
      "epoch": 0.00633,
      "grad_norm": 1.0593968691082751,
      "learning_rate": 0.001899,
      "loss": 5.1321,
      "step": 633
    },
    {
      "epoch": 0.00634,
      "grad_norm": 0.943443350801409,
      "learning_rate": 0.001902,
      "loss": 5.1182,
      "step": 634
    },
    {
      "epoch": 0.00635,
      "grad_norm": 0.9628484504873114,
      "learning_rate": 0.001905,
      "loss": 5.1089,
      "step": 635
    },
    {
      "epoch": 0.00636,
      "grad_norm": 1.1043383962750646,
      "learning_rate": 0.001908,
      "loss": 5.1289,
      "step": 636
    },
    {
      "epoch": 0.00637,
      "grad_norm": 0.8987493866500654,
      "learning_rate": 0.0019110000000000002,
      "loss": 5.1351,
      "step": 637
    },
    {
      "epoch": 0.00638,
      "grad_norm": 0.9251804377428581,
      "learning_rate": 0.0019140000000000001,
      "loss": 5.1288,
      "step": 638
    },
    {
      "epoch": 0.00639,
      "grad_norm": 0.854418425044198,
      "learning_rate": 0.001917,
      "loss": 5.0998,
      "step": 639
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.9324816679284724,
      "learning_rate": 0.00192,
      "loss": 5.1038,
      "step": 640
    },
    {
      "epoch": 0.00641,
      "grad_norm": 0.9892889234371413,
      "learning_rate": 0.001923,
      "loss": 5.1163,
      "step": 641
    },
    {
      "epoch": 0.00642,
      "grad_norm": 1.0346602459121752,
      "learning_rate": 0.001926,
      "loss": 5.106,
      "step": 642
    },
    {
      "epoch": 0.00643,
      "grad_norm": 0.8661994645957561,
      "learning_rate": 0.0019290000000000002,
      "loss": 5.117,
      "step": 643
    },
    {
      "epoch": 0.00644,
      "grad_norm": 0.8724056100423225,
      "learning_rate": 0.0019320000000000001,
      "loss": 5.0889,
      "step": 644
    },
    {
      "epoch": 0.00645,
      "grad_norm": 0.8584186184200229,
      "learning_rate": 0.001935,
      "loss": 5.1004,
      "step": 645
    },
    {
      "epoch": 0.00646,
      "grad_norm": 0.7360558672224548,
      "learning_rate": 0.001938,
      "loss": 5.0955,
      "step": 646
    },
    {
      "epoch": 0.00647,
      "grad_norm": 0.7977702647925389,
      "learning_rate": 0.001941,
      "loss": 5.1058,
      "step": 647
    },
    {
      "epoch": 0.00648,
      "grad_norm": 0.7872116543506851,
      "learning_rate": 0.0019440000000000002,
      "loss": 5.0908,
      "step": 648
    },
    {
      "epoch": 0.00649,
      "grad_norm": 0.7104658813349117,
      "learning_rate": 0.0019470000000000002,
      "loss": 5.0718,
      "step": 649
    },
    {
      "epoch": 0.0065,
      "grad_norm": 0.7453763255239747,
      "learning_rate": 0.0019500000000000001,
      "loss": 5.0953,
      "step": 650
    },
    {
      "epoch": 0.00651,
      "grad_norm": 0.7781624388594444,
      "learning_rate": 0.001953,
      "loss": 5.0758,
      "step": 651
    },
    {
      "epoch": 0.00652,
      "grad_norm": 0.7616046275009601,
      "learning_rate": 0.0019560000000000003,
      "loss": 5.0661,
      "step": 652
    },
    {
      "epoch": 0.00653,
      "grad_norm": 0.5945469625366651,
      "learning_rate": 0.0019590000000000002,
      "loss": 5.0539,
      "step": 653
    },
    {
      "epoch": 0.00654,
      "grad_norm": 0.6024408595794577,
      "learning_rate": 0.001962,
      "loss": 5.0374,
      "step": 654
    },
    {
      "epoch": 0.00655,
      "grad_norm": 0.5905307565923603,
      "learning_rate": 0.001965,
      "loss": 5.048,
      "step": 655
    },
    {
      "epoch": 0.00656,
      "grad_norm": 0.5236322372626927,
      "learning_rate": 0.001968,
      "loss": 5.04,
      "step": 656
    },
    {
      "epoch": 0.00657,
      "grad_norm": 0.5283416618835216,
      "learning_rate": 0.001971,
      "loss": 5.0223,
      "step": 657
    },
    {
      "epoch": 0.00658,
      "grad_norm": 0.5563146586062104,
      "learning_rate": 0.001974,
      "loss": 5.0415,
      "step": 658
    },
    {
      "epoch": 0.00659,
      "grad_norm": 0.6297873363395704,
      "learning_rate": 0.001977,
      "loss": 5.0241,
      "step": 659
    },
    {
      "epoch": 0.0066,
      "grad_norm": 0.5780538180580159,
      "learning_rate": 0.00198,
      "loss": 5.0197,
      "step": 660
    },
    {
      "epoch": 0.00661,
      "grad_norm": 0.5505266007864265,
      "learning_rate": 0.001983,
      "loss": 4.9938,
      "step": 661
    },
    {
      "epoch": 0.00662,
      "grad_norm": 0.6077995116545319,
      "learning_rate": 0.0019860000000000004,
      "loss": 4.9946,
      "step": 662
    },
    {
      "epoch": 0.00663,
      "grad_norm": 0.693544741809416,
      "learning_rate": 0.0019890000000000003,
      "loss": 5.01,
      "step": 663
    },
    {
      "epoch": 0.00664,
      "grad_norm": 0.801492715102265,
      "learning_rate": 0.0019920000000000003,
      "loss": 5.0017,
      "step": 664
    },
    {
      "epoch": 0.00665,
      "grad_norm": 0.8803866066519176,
      "learning_rate": 0.0019950000000000002,
      "loss": 4.9845,
      "step": 665
    },
    {
      "epoch": 0.00666,
      "grad_norm": 0.9133314823416234,
      "learning_rate": 0.001998,
      "loss": 5.0048,
      "step": 666
    },
    {
      "epoch": 0.00667,
      "grad_norm": 0.9006055773427947,
      "learning_rate": 0.002001,
      "loss": 4.9967,
      "step": 667
    },
    {
      "epoch": 0.00668,
      "grad_norm": 0.8268536663020751,
      "learning_rate": 0.002004,
      "loss": 5.0164,
      "step": 668
    },
    {
      "epoch": 0.00669,
      "grad_norm": 1.1034218430158187,
      "learning_rate": 0.002007,
      "loss": 5.0358,
      "step": 669
    },
    {
      "epoch": 0.0067,
      "grad_norm": 1.0710112933622913,
      "learning_rate": 0.00201,
      "loss": 5.0286,
      "step": 670
    },
    {
      "epoch": 0.00671,
      "grad_norm": 1.0384170566197124,
      "learning_rate": 0.002013,
      "loss": 5.0085,
      "step": 671
    },
    {
      "epoch": 0.00672,
      "grad_norm": 0.9247071512184438,
      "learning_rate": 0.002016,
      "loss": 4.9973,
      "step": 672
    },
    {
      "epoch": 0.00673,
      "grad_norm": 0.9905174678816935,
      "learning_rate": 0.002019,
      "loss": 5.0151,
      "step": 673
    },
    {
      "epoch": 0.00674,
      "grad_norm": 0.8930692065183647,
      "learning_rate": 0.0020220000000000004,
      "loss": 5.0014,
      "step": 674
    },
    {
      "epoch": 0.00675,
      "grad_norm": 1.0067908289229996,
      "learning_rate": 0.0020250000000000003,
      "loss": 5.0182,
      "step": 675
    },
    {
      "epoch": 0.00676,
      "grad_norm": 0.972209366764115,
      "learning_rate": 0.0020280000000000003,
      "loss": 5.0082,
      "step": 676
    },
    {
      "epoch": 0.00677,
      "grad_norm": 1.0440090424594235,
      "learning_rate": 0.0020310000000000003,
      "loss": 5.0109,
      "step": 677
    },
    {
      "epoch": 0.00678,
      "grad_norm": 1.1292207401563255,
      "learning_rate": 0.0020340000000000002,
      "loss": 5.0278,
      "step": 678
    },
    {
      "epoch": 0.00679,
      "grad_norm": 0.9872089241398577,
      "learning_rate": 0.002037,
      "loss": 5.0054,
      "step": 679
    },
    {
      "epoch": 0.0068,
      "grad_norm": 0.8748535789650627,
      "learning_rate": 0.00204,
      "loss": 5.0004,
      "step": 680
    },
    {
      "epoch": 0.00681,
      "grad_norm": 1.0039141852392888,
      "learning_rate": 0.002043,
      "loss": 5.0018,
      "step": 681
    },
    {
      "epoch": 0.00682,
      "grad_norm": 1.0414085118011525,
      "learning_rate": 0.002046,
      "loss": 5.0026,
      "step": 682
    },
    {
      "epoch": 0.00683,
      "grad_norm": 0.8083819733286619,
      "learning_rate": 0.002049,
      "loss": 4.9746,
      "step": 683
    },
    {
      "epoch": 0.00684,
      "grad_norm": 0.618682103661302,
      "learning_rate": 0.002052,
      "loss": 4.9803,
      "step": 684
    },
    {
      "epoch": 0.00685,
      "grad_norm": 0.5775508276061283,
      "learning_rate": 0.0020550000000000004,
      "loss": 4.9792,
      "step": 685
    },
    {
      "epoch": 0.00686,
      "grad_norm": 0.5359292821422027,
      "learning_rate": 0.0020580000000000004,
      "loss": 4.9508,
      "step": 686
    },
    {
      "epoch": 0.00687,
      "grad_norm": 0.5439874489161504,
      "learning_rate": 0.0020610000000000003,
      "loss": 4.9456,
      "step": 687
    },
    {
      "epoch": 0.00688,
      "grad_norm": 0.6521058673701751,
      "learning_rate": 0.002064,
      "loss": 4.9472,
      "step": 688
    },
    {
      "epoch": 0.00689,
      "grad_norm": 0.7201992210148584,
      "learning_rate": 0.002067,
      "loss": 4.9514,
      "step": 689
    },
    {
      "epoch": 0.0069,
      "grad_norm": 0.6329359289093167,
      "learning_rate": 0.00207,
      "loss": 4.9393,
      "step": 690
    },
    {
      "epoch": 0.00691,
      "grad_norm": 0.5452341441548267,
      "learning_rate": 0.0020729999999999998,
      "loss": 4.9065,
      "step": 691
    },
    {
      "epoch": 0.00692,
      "grad_norm": 0.524421270821715,
      "learning_rate": 0.0020759999999999997,
      "loss": 4.9264,
      "step": 692
    },
    {
      "epoch": 0.00693,
      "grad_norm": 0.5237198583423548,
      "learning_rate": 0.0020789999999999997,
      "loss": 4.9275,
      "step": 693
    },
    {
      "epoch": 0.00694,
      "grad_norm": 0.5943284166970615,
      "learning_rate": 0.002082,
      "loss": 4.9065,
      "step": 694
    },
    {
      "epoch": 0.00695,
      "grad_norm": 0.6331066584123409,
      "learning_rate": 0.002085,
      "loss": 4.9096,
      "step": 695
    },
    {
      "epoch": 0.00696,
      "grad_norm": 0.6552141599381052,
      "learning_rate": 0.002088,
      "loss": 4.9103,
      "step": 696
    },
    {
      "epoch": 0.00697,
      "grad_norm": 0.7459258980378775,
      "learning_rate": 0.002091,
      "loss": 4.9188,
      "step": 697
    },
    {
      "epoch": 0.00698,
      "grad_norm": 0.902413176094811,
      "learning_rate": 0.002094,
      "loss": 4.9109,
      "step": 698
    },
    {
      "epoch": 0.00699,
      "grad_norm": 0.9488326760888935,
      "learning_rate": 0.002097,
      "loss": 4.9116,
      "step": 699
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.9189763967394823,
      "learning_rate": 0.0021,
      "loss": 4.929,
      "step": 700
    },
    {
      "epoch": 0.00701,
      "grad_norm": 0.8924436046521577,
      "learning_rate": 0.002103,
      "loss": 4.9083,
      "step": 701
    },
    {
      "epoch": 0.00702,
      "grad_norm": 0.8863614629240012,
      "learning_rate": 0.002106,
      "loss": 4.9092,
      "step": 702
    },
    {
      "epoch": 0.00703,
      "grad_norm": 0.7940709541538681,
      "learning_rate": 0.0021089999999999998,
      "loss": 4.891,
      "step": 703
    },
    {
      "epoch": 0.00704,
      "grad_norm": 0.7939787736751149,
      "learning_rate": 0.0021119999999999997,
      "loss": 4.8964,
      "step": 704
    },
    {
      "epoch": 0.00705,
      "grad_norm": 1.0444656030359551,
      "learning_rate": 0.002115,
      "loss": 4.9151,
      "step": 705
    },
    {
      "epoch": 0.00706,
      "grad_norm": 1.1431311909042268,
      "learning_rate": 0.002118,
      "loss": 4.9234,
      "step": 706
    },
    {
      "epoch": 0.00707,
      "grad_norm": 0.8384635334186645,
      "learning_rate": 0.002121,
      "loss": 4.8902,
      "step": 707
    },
    {
      "epoch": 0.00708,
      "grad_norm": 0.9594405079672866,
      "learning_rate": 0.002124,
      "loss": 4.8745,
      "step": 708
    },
    {
      "epoch": 0.00709,
      "grad_norm": 0.8900382869322284,
      "learning_rate": 0.002127,
      "loss": 4.9161,
      "step": 709
    },
    {
      "epoch": 0.0071,
      "grad_norm": 0.8868570415283396,
      "learning_rate": 0.00213,
      "loss": 4.9065,
      "step": 710
    },
    {
      "epoch": 0.00711,
      "grad_norm": 0.8610490944817158,
      "learning_rate": 0.002133,
      "loss": 4.8804,
      "step": 711
    },
    {
      "epoch": 0.00712,
      "grad_norm": 0.8727164938852855,
      "learning_rate": 0.002136,
      "loss": 4.9046,
      "step": 712
    },
    {
      "epoch": 0.00713,
      "grad_norm": 0.8857525308493206,
      "learning_rate": 0.002139,
      "loss": 4.9135,
      "step": 713
    },
    {
      "epoch": 0.00714,
      "grad_norm": 0.9495661806955594,
      "learning_rate": 0.002142,
      "loss": 4.918,
      "step": 714
    },
    {
      "epoch": 0.00715,
      "grad_norm": 1.0263024097609161,
      "learning_rate": 0.0021449999999999998,
      "loss": 4.8857,
      "step": 715
    },
    {
      "epoch": 0.00716,
      "grad_norm": 0.8876358680026493,
      "learning_rate": 0.002148,
      "loss": 4.8749,
      "step": 716
    },
    {
      "epoch": 0.00717,
      "grad_norm": 0.8225498605776377,
      "learning_rate": 0.002151,
      "loss": 4.8925,
      "step": 717
    },
    {
      "epoch": 0.00718,
      "grad_norm": 0.628552485265691,
      "learning_rate": 0.002154,
      "loss": 4.8659,
      "step": 718
    },
    {
      "epoch": 0.00719,
      "grad_norm": 0.6584104654465238,
      "learning_rate": 0.002157,
      "loss": 4.8747,
      "step": 719
    },
    {
      "epoch": 0.0072,
      "grad_norm": 0.6698592474865601,
      "learning_rate": 0.00216,
      "loss": 4.8635,
      "step": 720
    },
    {
      "epoch": 0.00721,
      "grad_norm": 0.6673590176314685,
      "learning_rate": 0.002163,
      "loss": 4.8639,
      "step": 721
    },
    {
      "epoch": 0.00722,
      "grad_norm": 0.6674098225397388,
      "learning_rate": 0.002166,
      "loss": 4.8386,
      "step": 722
    },
    {
      "epoch": 0.00723,
      "grad_norm": 0.6090726175552883,
      "learning_rate": 0.002169,
      "loss": 4.8464,
      "step": 723
    },
    {
      "epoch": 0.00724,
      "grad_norm": 0.6325507361418539,
      "learning_rate": 0.002172,
      "loss": 4.8403,
      "step": 724
    },
    {
      "epoch": 0.00725,
      "grad_norm": 0.6927587431932604,
      "learning_rate": 0.002175,
      "loss": 4.8341,
      "step": 725
    },
    {
      "epoch": 0.00726,
      "grad_norm": 0.7422551683158218,
      "learning_rate": 0.002178,
      "loss": 4.8448,
      "step": 726
    },
    {
      "epoch": 0.00727,
      "grad_norm": 0.7946686392459241,
      "learning_rate": 0.0021809999999999998,
      "loss": 4.8264,
      "step": 727
    },
    {
      "epoch": 0.00728,
      "grad_norm": 0.651194780867581,
      "learning_rate": 0.002184,
      "loss": 4.8373,
      "step": 728
    },
    {
      "epoch": 0.00729,
      "grad_norm": 0.5507866158426874,
      "learning_rate": 0.002187,
      "loss": 4.8279,
      "step": 729
    },
    {
      "epoch": 0.0073,
      "grad_norm": 0.5770531279665235,
      "learning_rate": 0.00219,
      "loss": 4.8256,
      "step": 730
    },
    {
      "epoch": 0.00731,
      "grad_norm": 0.6604772562967653,
      "learning_rate": 0.002193,
      "loss": 4.8198,
      "step": 731
    },
    {
      "epoch": 0.00732,
      "grad_norm": 0.7902754963422924,
      "learning_rate": 0.002196,
      "loss": 4.832,
      "step": 732
    },
    {
      "epoch": 0.00733,
      "grad_norm": 0.8530754775219535,
      "learning_rate": 0.002199,
      "loss": 4.814,
      "step": 733
    },
    {
      "epoch": 0.00734,
      "grad_norm": 0.808835939559151,
      "learning_rate": 0.002202,
      "loss": 4.8365,
      "step": 734
    },
    {
      "epoch": 0.00735,
      "grad_norm": 0.7793455630355729,
      "learning_rate": 0.002205,
      "loss": 4.8484,
      "step": 735
    },
    {
      "epoch": 0.00736,
      "grad_norm": 0.8899384636665534,
      "learning_rate": 0.002208,
      "loss": 4.8322,
      "step": 736
    },
    {
      "epoch": 0.00737,
      "grad_norm": 1.0166382867407526,
      "learning_rate": 0.002211,
      "loss": 4.8145,
      "step": 737
    },
    {
      "epoch": 0.00738,
      "grad_norm": 0.9857126416807526,
      "learning_rate": 0.002214,
      "loss": 4.8213,
      "step": 738
    },
    {
      "epoch": 0.00739,
      "grad_norm": 0.8982016923721464,
      "learning_rate": 0.0022170000000000002,
      "loss": 4.809,
      "step": 739
    },
    {
      "epoch": 0.0074,
      "grad_norm": 0.8517808259905928,
      "learning_rate": 0.00222,
      "loss": 4.8138,
      "step": 740
    },
    {
      "epoch": 0.00741,
      "grad_norm": 0.6571636698582773,
      "learning_rate": 0.002223,
      "loss": 4.778,
      "step": 741
    },
    {
      "epoch": 0.00742,
      "grad_norm": 0.5983530041008951,
      "learning_rate": 0.002226,
      "loss": 4.8043,
      "step": 742
    },
    {
      "epoch": 0.00743,
      "grad_norm": 0.613767022264535,
      "learning_rate": 0.002229,
      "loss": 4.772,
      "step": 743
    },
    {
      "epoch": 0.00744,
      "grad_norm": 0.5746332772613801,
      "learning_rate": 0.002232,
      "loss": 4.7755,
      "step": 744
    },
    {
      "epoch": 0.00745,
      "grad_norm": 0.5491916195482714,
      "learning_rate": 0.002235,
      "loss": 4.7792,
      "step": 745
    },
    {
      "epoch": 0.00746,
      "grad_norm": 0.6224704559098753,
      "learning_rate": 0.002238,
      "loss": 4.7785,
      "step": 746
    },
    {
      "epoch": 0.00747,
      "grad_norm": 0.721423929849636,
      "learning_rate": 0.002241,
      "loss": 4.766,
      "step": 747
    },
    {
      "epoch": 0.00748,
      "grad_norm": 0.8211263973312402,
      "learning_rate": 0.002244,
      "loss": 4.7825,
      "step": 748
    },
    {
      "epoch": 0.00749,
      "grad_norm": 0.9630311220772746,
      "learning_rate": 0.002247,
      "loss": 4.7687,
      "step": 749
    },
    {
      "epoch": 0.0075,
      "grad_norm": 1.0671208539128567,
      "learning_rate": 0.0022500000000000003,
      "loss": 4.8201,
      "step": 750
    },
    {
      "epoch": 0.00751,
      "grad_norm": 1.0285377005887373,
      "learning_rate": 0.0022530000000000002,
      "loss": 4.8053,
      "step": 751
    },
    {
      "epoch": 0.00752,
      "grad_norm": 0.9937387135055332,
      "learning_rate": 0.002256,
      "loss": 4.7635,
      "step": 752
    },
    {
      "epoch": 0.00753,
      "grad_norm": 0.8939099354397223,
      "learning_rate": 0.002259,
      "loss": 4.8048,
      "step": 753
    },
    {
      "epoch": 0.00754,
      "grad_norm": 0.958657384547811,
      "learning_rate": 0.002262,
      "loss": 4.816,
      "step": 754
    },
    {
      "epoch": 0.00755,
      "grad_norm": 0.8579165829387244,
      "learning_rate": 0.002265,
      "loss": 4.7809,
      "step": 755
    },
    {
      "epoch": 0.00756,
      "grad_norm": 0.8147493051985796,
      "learning_rate": 0.002268,
      "loss": 4.7687,
      "step": 756
    },
    {
      "epoch": 0.00757,
      "grad_norm": 0.9899674342411924,
      "learning_rate": 0.002271,
      "loss": 4.7845,
      "step": 757
    },
    {
      "epoch": 0.00758,
      "grad_norm": 1.2327208130150207,
      "learning_rate": 0.002274,
      "loss": 4.7967,
      "step": 758
    },
    {
      "epoch": 0.00759,
      "grad_norm": 0.8760473410677282,
      "learning_rate": 0.002277,
      "loss": 4.7951,
      "step": 759
    },
    {
      "epoch": 0.0076,
      "grad_norm": 0.9553393765090987,
      "learning_rate": 0.00228,
      "loss": 4.7842,
      "step": 760
    },
    {
      "epoch": 0.00761,
      "grad_norm": 1.088855928225056,
      "learning_rate": 0.002283,
      "loss": 4.769,
      "step": 761
    },
    {
      "epoch": 0.00762,
      "grad_norm": 0.9818480683263884,
      "learning_rate": 0.0022860000000000003,
      "loss": 4.7512,
      "step": 762
    },
    {
      "epoch": 0.00763,
      "grad_norm": 0.9731540924632093,
      "learning_rate": 0.0022890000000000002,
      "loss": 4.7931,
      "step": 763
    },
    {
      "epoch": 0.00764,
      "grad_norm": 1.0508884929557651,
      "learning_rate": 0.002292,
      "loss": 4.8167,
      "step": 764
    },
    {
      "epoch": 0.00765,
      "grad_norm": 1.0020863769727308,
      "learning_rate": 0.002295,
      "loss": 4.7984,
      "step": 765
    },
    {
      "epoch": 0.00766,
      "grad_norm": 1.1527463652354557,
      "learning_rate": 0.002298,
      "loss": 4.8085,
      "step": 766
    },
    {
      "epoch": 0.00767,
      "grad_norm": 0.9657952239159258,
      "learning_rate": 0.002301,
      "loss": 4.7959,
      "step": 767
    },
    {
      "epoch": 0.00768,
      "grad_norm": 1.0234917976922082,
      "learning_rate": 0.002304,
      "loss": 4.8012,
      "step": 768
    },
    {
      "epoch": 0.00769,
      "grad_norm": 0.9850893067060651,
      "learning_rate": 0.002307,
      "loss": 4.8144,
      "step": 769
    },
    {
      "epoch": 0.0077,
      "grad_norm": 0.9062134932024389,
      "learning_rate": 0.00231,
      "loss": 4.7653,
      "step": 770
    },
    {
      "epoch": 0.00771,
      "grad_norm": 0.8476285286232204,
      "learning_rate": 0.002313,
      "loss": 4.7979,
      "step": 771
    },
    {
      "epoch": 0.00772,
      "grad_norm": 0.9122213123018311,
      "learning_rate": 0.002316,
      "loss": 4.7851,
      "step": 772
    },
    {
      "epoch": 0.00773,
      "grad_norm": 1.0718910624781612,
      "learning_rate": 0.0023190000000000003,
      "loss": 4.8052,
      "step": 773
    },
    {
      "epoch": 0.00774,
      "grad_norm": 0.7792131883523417,
      "learning_rate": 0.0023220000000000003,
      "loss": 4.7945,
      "step": 774
    },
    {
      "epoch": 0.00775,
      "grad_norm": 0.7995411986928386,
      "learning_rate": 0.0023250000000000002,
      "loss": 4.7914,
      "step": 775
    },
    {
      "epoch": 0.00776,
      "grad_norm": 0.7054590225014301,
      "learning_rate": 0.002328,
      "loss": 4.7883,
      "step": 776
    },
    {
      "epoch": 0.00777,
      "grad_norm": 0.6505869359405926,
      "learning_rate": 0.002331,
      "loss": 4.7585,
      "step": 777
    },
    {
      "epoch": 0.00778,
      "grad_norm": 0.6484695284206986,
      "learning_rate": 0.002334,
      "loss": 4.7652,
      "step": 778
    },
    {
      "epoch": 0.00779,
      "grad_norm": 0.6047799586124271,
      "learning_rate": 0.002337,
      "loss": 4.7239,
      "step": 779
    },
    {
      "epoch": 0.0078,
      "grad_norm": 0.5436502526586032,
      "learning_rate": 0.00234,
      "loss": 4.7364,
      "step": 780
    },
    {
      "epoch": 0.00781,
      "grad_norm": 0.5682167623371829,
      "learning_rate": 0.002343,
      "loss": 4.7355,
      "step": 781
    },
    {
      "epoch": 0.00782,
      "grad_norm": 0.628910610744215,
      "learning_rate": 0.002346,
      "loss": 4.7356,
      "step": 782
    },
    {
      "epoch": 0.00783,
      "grad_norm": 0.5845457532196663,
      "learning_rate": 0.002349,
      "loss": 4.7086,
      "step": 783
    },
    {
      "epoch": 0.00784,
      "grad_norm": 0.6345965495249546,
      "learning_rate": 0.002352,
      "loss": 4.723,
      "step": 784
    },
    {
      "epoch": 0.00785,
      "grad_norm": 0.4959017064429773,
      "learning_rate": 0.0023550000000000003,
      "loss": 4.7138,
      "step": 785
    },
    {
      "epoch": 0.00786,
      "grad_norm": 0.4358915945164792,
      "learning_rate": 0.0023580000000000003,
      "loss": 4.7131,
      "step": 786
    },
    {
      "epoch": 0.00787,
      "grad_norm": 0.4259891850743534,
      "learning_rate": 0.0023610000000000003,
      "loss": 4.7113,
      "step": 787
    },
    {
      "epoch": 0.00788,
      "grad_norm": 0.4529949912379573,
      "learning_rate": 0.002364,
      "loss": 4.7077,
      "step": 788
    },
    {
      "epoch": 0.00789,
      "grad_norm": 0.4944753699230628,
      "learning_rate": 0.002367,
      "loss": 4.7334,
      "step": 789
    },
    {
      "epoch": 0.0079,
      "grad_norm": 0.6295031827770177,
      "learning_rate": 0.00237,
      "loss": 4.7144,
      "step": 790
    },
    {
      "epoch": 0.00791,
      "grad_norm": 0.8474014146600959,
      "learning_rate": 0.002373,
      "loss": 4.6777,
      "step": 791
    },
    {
      "epoch": 0.00792,
      "grad_norm": 0.8040402155973354,
      "learning_rate": 0.002376,
      "loss": 4.7267,
      "step": 792
    },
    {
      "epoch": 0.00793,
      "grad_norm": 0.5568568781614048,
      "learning_rate": 0.002379,
      "loss": 4.7006,
      "step": 793
    },
    {
      "epoch": 0.00794,
      "grad_norm": 0.8220014797505664,
      "learning_rate": 0.002382,
      "loss": 4.7246,
      "step": 794
    },
    {
      "epoch": 0.00795,
      "grad_norm": 0.6551332917875898,
      "learning_rate": 0.002385,
      "loss": 4.6938,
      "step": 795
    },
    {
      "epoch": 0.00796,
      "grad_norm": 0.5666814801389223,
      "learning_rate": 0.0023880000000000004,
      "loss": 4.6826,
      "step": 796
    },
    {
      "epoch": 0.00797,
      "grad_norm": 0.654430341218369,
      "learning_rate": 0.0023910000000000003,
      "loss": 4.7248,
      "step": 797
    },
    {
      "epoch": 0.00798,
      "grad_norm": 0.5511512463730408,
      "learning_rate": 0.0023940000000000003,
      "loss": 4.6864,
      "step": 798
    },
    {
      "epoch": 0.00799,
      "grad_norm": 0.5084174359945534,
      "learning_rate": 0.0023970000000000003,
      "loss": 4.7161,
      "step": 799
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.41968398472595075,
      "learning_rate": 0.0024000000000000002,
      "loss": 4.6652,
      "step": 800
    },
    {
      "epoch": 0.00801,
      "grad_norm": 0.4546309911468048,
      "learning_rate": 0.002403,
      "loss": 4.6967,
      "step": 801
    },
    {
      "epoch": 0.00802,
      "grad_norm": 0.430018994115786,
      "learning_rate": 0.002406,
      "loss": 4.6827,
      "step": 802
    },
    {
      "epoch": 0.00803,
      "grad_norm": 0.44614830716967085,
      "learning_rate": 0.002409,
      "loss": 4.6617,
      "step": 803
    },
    {
      "epoch": 0.00804,
      "grad_norm": 0.44722400844593674,
      "learning_rate": 0.002412,
      "loss": 4.6598,
      "step": 804
    },
    {
      "epoch": 0.00805,
      "grad_norm": 0.5179127215582825,
      "learning_rate": 0.002415,
      "loss": 4.6599,
      "step": 805
    },
    {
      "epoch": 0.00806,
      "grad_norm": 0.5610832008078775,
      "learning_rate": 0.002418,
      "loss": 4.677,
      "step": 806
    },
    {
      "epoch": 0.00807,
      "grad_norm": 0.5167453223410896,
      "learning_rate": 0.0024210000000000004,
      "loss": 4.6671,
      "step": 807
    },
    {
      "epoch": 0.00808,
      "grad_norm": 0.46468933196331563,
      "learning_rate": 0.0024240000000000004,
      "loss": 4.6511,
      "step": 808
    },
    {
      "epoch": 0.00809,
      "grad_norm": 0.5221883532574668,
      "learning_rate": 0.0024270000000000003,
      "loss": 4.6468,
      "step": 809
    },
    {
      "epoch": 0.0081,
      "grad_norm": 0.4992566900849729,
      "learning_rate": 0.0024300000000000003,
      "loss": 4.6744,
      "step": 810
    },
    {
      "epoch": 0.00811,
      "grad_norm": 0.4854147467055134,
      "learning_rate": 0.0024330000000000003,
      "loss": 4.646,
      "step": 811
    },
    {
      "epoch": 0.00812,
      "grad_norm": 0.650970729431075,
      "learning_rate": 0.0024360000000000002,
      "loss": 4.6307,
      "step": 812
    },
    {
      "epoch": 0.00813,
      "grad_norm": 0.8160691589494683,
      "learning_rate": 0.0024389999999999998,
      "loss": 4.6711,
      "step": 813
    },
    {
      "epoch": 0.00814,
      "grad_norm": 0.9918101747931352,
      "learning_rate": 0.0024419999999999997,
      "loss": 4.6946,
      "step": 814
    },
    {
      "epoch": 0.00815,
      "grad_norm": 1.247963175893729,
      "learning_rate": 0.0024449999999999997,
      "loss": 4.7226,
      "step": 815
    },
    {
      "epoch": 0.00816,
      "grad_norm": 0.8376200515557375,
      "learning_rate": 0.002448,
      "loss": 4.6777,
      "step": 816
    },
    {
      "epoch": 0.00817,
      "grad_norm": 0.9161032619759178,
      "learning_rate": 0.002451,
      "loss": 4.6939,
      "step": 817
    },
    {
      "epoch": 0.00818,
      "grad_norm": 1.0914649908014256,
      "learning_rate": 0.002454,
      "loss": 4.6886,
      "step": 818
    },
    {
      "epoch": 0.00819,
      "grad_norm": 0.9806171410774952,
      "learning_rate": 0.002457,
      "loss": 4.712,
      "step": 819
    },
    {
      "epoch": 0.0082,
      "grad_norm": 0.992236077471004,
      "learning_rate": 0.00246,
      "loss": 4.6918,
      "step": 820
    },
    {
      "epoch": 0.00821,
      "grad_norm": 1.0594557870263281,
      "learning_rate": 0.002463,
      "loss": 4.6759,
      "step": 821
    },
    {
      "epoch": 0.00822,
      "grad_norm": 1.0346800919438124,
      "learning_rate": 0.002466,
      "loss": 4.6853,
      "step": 822
    },
    {
      "epoch": 0.00823,
      "grad_norm": 0.9573573191186882,
      "learning_rate": 0.002469,
      "loss": 4.6833,
      "step": 823
    },
    {
      "epoch": 0.00824,
      "grad_norm": 1.1123514933123841,
      "learning_rate": 0.002472,
      "loss": 4.714,
      "step": 824
    },
    {
      "epoch": 0.00825,
      "grad_norm": 0.8463845700248506,
      "learning_rate": 0.0024749999999999998,
      "loss": 4.7191,
      "step": 825
    },
    {
      "epoch": 0.00826,
      "grad_norm": 0.8444785606085857,
      "learning_rate": 0.0024779999999999997,
      "loss": 4.672,
      "step": 826
    },
    {
      "epoch": 0.00827,
      "grad_norm": 0.9726341870117121,
      "learning_rate": 0.002481,
      "loss": 4.7078,
      "step": 827
    },
    {
      "epoch": 0.00828,
      "grad_norm": 0.9106448417621353,
      "learning_rate": 0.002484,
      "loss": 4.7003,
      "step": 828
    },
    {
      "epoch": 0.00829,
      "grad_norm": 0.7565680418878746,
      "learning_rate": 0.002487,
      "loss": 4.6856,
      "step": 829
    },
    {
      "epoch": 0.0083,
      "grad_norm": 0.8537774465977133,
      "learning_rate": 0.00249,
      "loss": 4.7017,
      "step": 830
    },
    {
      "epoch": 0.00831,
      "grad_norm": 0.9023323948099834,
      "learning_rate": 0.002493,
      "loss": 4.6871,
      "step": 831
    },
    {
      "epoch": 0.00832,
      "grad_norm": 0.8524529451127855,
      "learning_rate": 0.002496,
      "loss": 4.6815,
      "step": 832
    },
    {
      "epoch": 0.00833,
      "grad_norm": 0.9428655185832147,
      "learning_rate": 0.002499,
      "loss": 4.6808,
      "step": 833
    },
    {
      "epoch": 0.00834,
      "grad_norm": 0.9597220185428569,
      "learning_rate": 0.002502,
      "loss": 4.6879,
      "step": 834
    },
    {
      "epoch": 0.00835,
      "grad_norm": 0.7735101632354252,
      "learning_rate": 0.002505,
      "loss": 4.6724,
      "step": 835
    },
    {
      "epoch": 0.00836,
      "grad_norm": 0.9597202731139803,
      "learning_rate": 0.002508,
      "loss": 4.7061,
      "step": 836
    },
    {
      "epoch": 0.00837,
      "grad_norm": 0.9520863539431935,
      "learning_rate": 0.0025109999999999998,
      "loss": 4.6636,
      "step": 837
    },
    {
      "epoch": 0.00838,
      "grad_norm": 0.7800128524395746,
      "learning_rate": 0.0025139999999999997,
      "loss": 4.6721,
      "step": 838
    },
    {
      "epoch": 0.00839,
      "grad_norm": 0.8122589832425033,
      "learning_rate": 0.002517,
      "loss": 4.675,
      "step": 839
    },
    {
      "epoch": 0.0084,
      "grad_norm": 0.8183344402395425,
      "learning_rate": 0.00252,
      "loss": 4.6669,
      "step": 840
    },
    {
      "epoch": 0.00841,
      "grad_norm": 0.6918735110390536,
      "learning_rate": 0.002523,
      "loss": 4.6489,
      "step": 841
    },
    {
      "epoch": 0.00842,
      "grad_norm": 0.6201385747244391,
      "learning_rate": 0.002526,
      "loss": 4.6423,
      "step": 842
    },
    {
      "epoch": 0.00843,
      "grad_norm": 0.606127970479136,
      "learning_rate": 0.002529,
      "loss": 4.6465,
      "step": 843
    },
    {
      "epoch": 0.00844,
      "grad_norm": 0.5515773209874846,
      "learning_rate": 0.002532,
      "loss": 4.6607,
      "step": 844
    },
    {
      "epoch": 0.00845,
      "grad_norm": 0.6203742299859808,
      "learning_rate": 0.002535,
      "loss": 4.6293,
      "step": 845
    },
    {
      "epoch": 0.00846,
      "grad_norm": 0.5875832865020281,
      "learning_rate": 0.002538,
      "loss": 4.6474,
      "step": 846
    },
    {
      "epoch": 0.00847,
      "grad_norm": 0.5703256353430879,
      "learning_rate": 0.002541,
      "loss": 4.6282,
      "step": 847
    },
    {
      "epoch": 0.00848,
      "grad_norm": 0.602830367643936,
      "learning_rate": 0.002544,
      "loss": 4.6269,
      "step": 848
    },
    {
      "epoch": 0.00849,
      "grad_norm": 0.6741507039909044,
      "learning_rate": 0.002547,
      "loss": 4.6233,
      "step": 849
    },
    {
      "epoch": 0.0085,
      "grad_norm": 0.6288739006540759,
      "learning_rate": 0.00255,
      "loss": 4.6341,
      "step": 850
    },
    {
      "epoch": 0.00851,
      "grad_norm": 0.5820099008678455,
      "learning_rate": 0.002553,
      "loss": 4.644,
      "step": 851
    },
    {
      "epoch": 0.00852,
      "grad_norm": 0.586123912558797,
      "learning_rate": 0.002556,
      "loss": 4.6367,
      "step": 852
    },
    {
      "epoch": 0.00853,
      "grad_norm": 0.5127813487098001,
      "learning_rate": 0.002559,
      "loss": 4.6085,
      "step": 853
    },
    {
      "epoch": 0.00854,
      "grad_norm": 0.4730499644759358,
      "learning_rate": 0.002562,
      "loss": 4.6029,
      "step": 854
    },
    {
      "epoch": 0.00855,
      "grad_norm": 0.44708869980986227,
      "learning_rate": 0.002565,
      "loss": 4.5799,
      "step": 855
    },
    {
      "epoch": 0.00856,
      "grad_norm": 0.466044480858233,
      "learning_rate": 0.002568,
      "loss": 4.6142,
      "step": 856
    },
    {
      "epoch": 0.00857,
      "grad_norm": 0.5382201915945353,
      "learning_rate": 0.002571,
      "loss": 4.6036,
      "step": 857
    },
    {
      "epoch": 0.00858,
      "grad_norm": 0.6780662034295477,
      "learning_rate": 0.002574,
      "loss": 4.609,
      "step": 858
    },
    {
      "epoch": 0.00859,
      "grad_norm": 0.9086610382483981,
      "learning_rate": 0.002577,
      "loss": 4.6039,
      "step": 859
    },
    {
      "epoch": 0.0086,
      "grad_norm": 0.8563688949272525,
      "learning_rate": 0.00258,
      "loss": 4.6531,
      "step": 860
    },
    {
      "epoch": 0.00861,
      "grad_norm": 0.5965670098126366,
      "learning_rate": 0.0025830000000000002,
      "loss": 4.6461,
      "step": 861
    },
    {
      "epoch": 0.00862,
      "grad_norm": 0.7975052365958228,
      "learning_rate": 0.002586,
      "loss": 4.611,
      "step": 862
    },
    {
      "epoch": 0.00863,
      "grad_norm": 0.650099032572018,
      "learning_rate": 0.002589,
      "loss": 4.6284,
      "step": 863
    },
    {
      "epoch": 0.00864,
      "grad_norm": 0.6277114763068243,
      "learning_rate": 0.002592,
      "loss": 4.5809,
      "step": 864
    },
    {
      "epoch": 0.00865,
      "grad_norm": 0.7499269309750987,
      "learning_rate": 0.002595,
      "loss": 4.6024,
      "step": 865
    },
    {
      "epoch": 0.00866,
      "grad_norm": 0.7085973518403954,
      "learning_rate": 0.002598,
      "loss": 4.6233,
      "step": 866
    },
    {
      "epoch": 0.00867,
      "grad_norm": 0.605538925445329,
      "learning_rate": 0.002601,
      "loss": 4.6101,
      "step": 867
    },
    {
      "epoch": 0.00868,
      "grad_norm": 0.48200776305054654,
      "learning_rate": 0.002604,
      "loss": 4.586,
      "step": 868
    },
    {
      "epoch": 0.00869,
      "grad_norm": 0.5266950965425763,
      "learning_rate": 0.002607,
      "loss": 4.5966,
      "step": 869
    },
    {
      "epoch": 0.0087,
      "grad_norm": 0.48953699231607295,
      "learning_rate": 0.00261,
      "loss": 4.5832,
      "step": 870
    },
    {
      "epoch": 0.00871,
      "grad_norm": 0.5478274928438833,
      "learning_rate": 0.002613,
      "loss": 4.5992,
      "step": 871
    },
    {
      "epoch": 0.00872,
      "grad_norm": 0.6263670684952429,
      "learning_rate": 0.002616,
      "loss": 4.5864,
      "step": 872
    },
    {
      "epoch": 0.00873,
      "grad_norm": 0.6672951132542,
      "learning_rate": 0.0026190000000000002,
      "loss": 4.5977,
      "step": 873
    },
    {
      "epoch": 0.00874,
      "grad_norm": 0.677096167715366,
      "learning_rate": 0.002622,
      "loss": 4.617,
      "step": 874
    },
    {
      "epoch": 0.00875,
      "grad_norm": 0.6959913524482387,
      "learning_rate": 0.002625,
      "loss": 4.5696,
      "step": 875
    },
    {
      "epoch": 0.00876,
      "grad_norm": 0.6682762743495083,
      "learning_rate": 0.002628,
      "loss": 4.5958,
      "step": 876
    },
    {
      "epoch": 0.00877,
      "grad_norm": 0.6375266502117092,
      "learning_rate": 0.002631,
      "loss": 4.612,
      "step": 877
    },
    {
      "epoch": 0.00878,
      "grad_norm": 0.7079418482290942,
      "learning_rate": 0.002634,
      "loss": 4.5486,
      "step": 878
    },
    {
      "epoch": 0.00879,
      "grad_norm": 0.6282689223941402,
      "learning_rate": 0.002637,
      "loss": 4.578,
      "step": 879
    },
    {
      "epoch": 0.0088,
      "grad_norm": 0.5460943252882049,
      "learning_rate": 0.00264,
      "loss": 4.5852,
      "step": 880
    },
    {
      "epoch": 0.00881,
      "grad_norm": 0.5723972494402886,
      "learning_rate": 0.002643,
      "loss": 4.5869,
      "step": 881
    },
    {
      "epoch": 0.00882,
      "grad_norm": 0.647818443655113,
      "learning_rate": 0.002646,
      "loss": 4.588,
      "step": 882
    },
    {
      "epoch": 0.00883,
      "grad_norm": 0.8827068805337381,
      "learning_rate": 0.002649,
      "loss": 4.5935,
      "step": 883
    },
    {
      "epoch": 0.00884,
      "grad_norm": 1.3000201706023533,
      "learning_rate": 0.0026520000000000003,
      "loss": 4.6052,
      "step": 884
    },
    {
      "epoch": 0.00885,
      "grad_norm": 0.7527768384442359,
      "learning_rate": 0.0026550000000000002,
      "loss": 4.5797,
      "step": 885
    },
    {
      "epoch": 0.00886,
      "grad_norm": 0.7313595200920677,
      "learning_rate": 0.002658,
      "loss": 4.6019,
      "step": 886
    },
    {
      "epoch": 0.00887,
      "grad_norm": 0.553327654847044,
      "learning_rate": 0.002661,
      "loss": 4.5828,
      "step": 887
    },
    {
      "epoch": 0.00888,
      "grad_norm": 0.6064219625843388,
      "learning_rate": 0.002664,
      "loss": 4.5894,
      "step": 888
    },
    {
      "epoch": 0.00889,
      "grad_norm": 0.6392357596846293,
      "learning_rate": 0.002667,
      "loss": 4.5422,
      "step": 889
    },
    {
      "epoch": 0.0089,
      "grad_norm": 0.6860123914477424,
      "learning_rate": 0.00267,
      "loss": 4.5989,
      "step": 890
    },
    {
      "epoch": 0.00891,
      "grad_norm": 0.7088960904364014,
      "learning_rate": 0.002673,
      "loss": 4.5822,
      "step": 891
    },
    {
      "epoch": 0.00892,
      "grad_norm": 0.7157207147763361,
      "learning_rate": 0.002676,
      "loss": 4.5934,
      "step": 892
    },
    {
      "epoch": 0.00893,
      "grad_norm": 0.7412527752908875,
      "learning_rate": 0.002679,
      "loss": 4.5709,
      "step": 893
    },
    {
      "epoch": 0.00894,
      "grad_norm": 0.8084836835989728,
      "learning_rate": 0.002682,
      "loss": 4.5639,
      "step": 894
    },
    {
      "epoch": 0.00895,
      "grad_norm": 0.9923307111818513,
      "learning_rate": 0.0026850000000000003,
      "loss": 4.5864,
      "step": 895
    },
    {
      "epoch": 0.00896,
      "grad_norm": 1.2171682577354312,
      "learning_rate": 0.0026880000000000003,
      "loss": 4.6057,
      "step": 896
    },
    {
      "epoch": 0.00897,
      "grad_norm": 0.797478427208377,
      "learning_rate": 0.0026910000000000002,
      "loss": 4.5989,
      "step": 897
    },
    {
      "epoch": 0.00898,
      "grad_norm": 0.7928728804117916,
      "learning_rate": 0.002694,
      "loss": 4.594,
      "step": 898
    },
    {
      "epoch": 0.00899,
      "grad_norm": 0.8357403035452178,
      "learning_rate": 0.002697,
      "loss": 4.5983,
      "step": 899
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.8448290091163538,
      "learning_rate": 0.0027,
      "loss": 4.6292,
      "step": 900
    },
    {
      "epoch": 0.00901,
      "grad_norm": 0.9488092229670547,
      "learning_rate": 0.002703,
      "loss": 4.5868,
      "step": 901
    },
    {
      "epoch": 0.00902,
      "grad_norm": 0.9434404658743749,
      "learning_rate": 0.002706,
      "loss": 4.5999,
      "step": 902
    },
    {
      "epoch": 0.00903,
      "grad_norm": 1.0122099567822476,
      "learning_rate": 0.002709,
      "loss": 4.6102,
      "step": 903
    },
    {
      "epoch": 0.00904,
      "grad_norm": 0.9358691681287052,
      "learning_rate": 0.002712,
      "loss": 4.5848,
      "step": 904
    },
    {
      "epoch": 0.00905,
      "grad_norm": 0.8321510442485943,
      "learning_rate": 0.002715,
      "loss": 4.5984,
      "step": 905
    },
    {
      "epoch": 0.00906,
      "grad_norm": 0.8914473393947665,
      "learning_rate": 0.002718,
      "loss": 4.6112,
      "step": 906
    },
    {
      "epoch": 0.00907,
      "grad_norm": 0.9883982303638487,
      "learning_rate": 0.0027210000000000003,
      "loss": 4.6386,
      "step": 907
    },
    {
      "epoch": 0.00908,
      "grad_norm": 0.86073203349026,
      "learning_rate": 0.0027240000000000003,
      "loss": 4.6116,
      "step": 908
    },
    {
      "epoch": 0.00909,
      "grad_norm": 0.7773747412069614,
      "learning_rate": 0.0027270000000000003,
      "loss": 4.6163,
      "step": 909
    },
    {
      "epoch": 0.0091,
      "grad_norm": 0.7370585718531062,
      "learning_rate": 0.0027300000000000002,
      "loss": 4.6234,
      "step": 910
    },
    {
      "epoch": 0.00911,
      "grad_norm": 0.6906269071273593,
      "learning_rate": 0.002733,
      "loss": 4.5785,
      "step": 911
    },
    {
      "epoch": 0.00912,
      "grad_norm": 0.6578032292778252,
      "learning_rate": 0.002736,
      "loss": 4.5778,
      "step": 912
    },
    {
      "epoch": 0.00913,
      "grad_norm": 0.6528626059582382,
      "learning_rate": 0.002739,
      "loss": 4.5704,
      "step": 913
    },
    {
      "epoch": 0.00914,
      "grad_norm": 0.599731896856576,
      "learning_rate": 0.002742,
      "loss": 4.595,
      "step": 914
    },
    {
      "epoch": 0.00915,
      "grad_norm": 0.5922054086035364,
      "learning_rate": 0.002745,
      "loss": 4.5555,
      "step": 915
    },
    {
      "epoch": 0.00916,
      "grad_norm": 0.528646140228931,
      "learning_rate": 0.002748,
      "loss": 4.5304,
      "step": 916
    },
    {
      "epoch": 0.00917,
      "grad_norm": 0.5305158198561161,
      "learning_rate": 0.002751,
      "loss": 4.5419,
      "step": 917
    },
    {
      "epoch": 0.00918,
      "grad_norm": 0.4736382884122071,
      "learning_rate": 0.0027540000000000004,
      "loss": 4.5569,
      "step": 918
    },
    {
      "epoch": 0.00919,
      "grad_norm": 0.45838817911808083,
      "learning_rate": 0.0027570000000000003,
      "loss": 4.5357,
      "step": 919
    },
    {
      "epoch": 0.0092,
      "grad_norm": 0.4361472184478695,
      "learning_rate": 0.0027600000000000003,
      "loss": 4.5493,
      "step": 920
    },
    {
      "epoch": 0.00921,
      "grad_norm": 0.43931126757575867,
      "learning_rate": 0.0027630000000000003,
      "loss": 4.5336,
      "step": 921
    },
    {
      "epoch": 0.00922,
      "grad_norm": 0.46920285200840567,
      "learning_rate": 0.0027660000000000002,
      "loss": 4.5412,
      "step": 922
    },
    {
      "epoch": 0.00923,
      "grad_norm": 0.5996209745093354,
      "learning_rate": 0.002769,
      "loss": 4.5218,
      "step": 923
    },
    {
      "epoch": 0.00924,
      "grad_norm": 0.702528530852281,
      "learning_rate": 0.002772,
      "loss": 4.5564,
      "step": 924
    },
    {
      "epoch": 0.00925,
      "grad_norm": 0.6651779892024297,
      "learning_rate": 0.002775,
      "loss": 4.5288,
      "step": 925
    },
    {
      "epoch": 0.00926,
      "grad_norm": 0.5723855300218565,
      "learning_rate": 0.002778,
      "loss": 4.5426,
      "step": 926
    },
    {
      "epoch": 0.00927,
      "grad_norm": 0.6080148945846373,
      "learning_rate": 0.002781,
      "loss": 4.5043,
      "step": 927
    },
    {
      "epoch": 0.00928,
      "grad_norm": 0.6957035289314476,
      "learning_rate": 0.002784,
      "loss": 4.5502,
      "step": 928
    },
    {
      "epoch": 0.00929,
      "grad_norm": 0.7366634335117053,
      "learning_rate": 0.0027870000000000004,
      "loss": 4.5403,
      "step": 929
    },
    {
      "epoch": 0.0093,
      "grad_norm": 0.7796102692352119,
      "learning_rate": 0.0027900000000000004,
      "loss": 4.5265,
      "step": 930
    },
    {
      "epoch": 0.00931,
      "grad_norm": 0.7116909602884058,
      "learning_rate": 0.0027930000000000003,
      "loss": 4.5333,
      "step": 931
    },
    {
      "epoch": 0.00932,
      "grad_norm": 0.8319854024158545,
      "learning_rate": 0.0027960000000000003,
      "loss": 4.5481,
      "step": 932
    },
    {
      "epoch": 0.00933,
      "grad_norm": 0.8001418279766108,
      "learning_rate": 0.0027990000000000003,
      "loss": 4.5286,
      "step": 933
    },
    {
      "epoch": 0.00934,
      "grad_norm": 0.6519619232143173,
      "learning_rate": 0.0028020000000000002,
      "loss": 4.5422,
      "step": 934
    },
    {
      "epoch": 0.00935,
      "grad_norm": 0.7580737482550882,
      "learning_rate": 0.002805,
      "loss": 4.5559,
      "step": 935
    },
    {
      "epoch": 0.00936,
      "grad_norm": 0.8499038627491867,
      "learning_rate": 0.002808,
      "loss": 4.5679,
      "step": 936
    },
    {
      "epoch": 0.00937,
      "grad_norm": 0.7597302495348821,
      "learning_rate": 0.002811,
      "loss": 4.5708,
      "step": 937
    },
    {
      "epoch": 0.00938,
      "grad_norm": 0.9878821641788273,
      "learning_rate": 0.002814,
      "loss": 4.5609,
      "step": 938
    },
    {
      "epoch": 0.00939,
      "grad_norm": 0.9691729918808772,
      "learning_rate": 0.002817,
      "loss": 4.563,
      "step": 939
    },
    {
      "epoch": 0.0094,
      "grad_norm": 0.8937843559478598,
      "learning_rate": 0.00282,
      "loss": 4.56,
      "step": 940
    },
    {
      "epoch": 0.00941,
      "grad_norm": 0.9477839045288606,
      "learning_rate": 0.002823,
      "loss": 4.5409,
      "step": 941
    },
    {
      "epoch": 0.00942,
      "grad_norm": 1.088560613057821,
      "learning_rate": 0.002826,
      "loss": 4.5819,
      "step": 942
    },
    {
      "epoch": 0.00943,
      "grad_norm": 0.8020128186220904,
      "learning_rate": 0.002829,
      "loss": 4.556,
      "step": 943
    },
    {
      "epoch": 0.00944,
      "grad_norm": 0.7970499406732843,
      "learning_rate": 0.002832,
      "loss": 4.5652,
      "step": 944
    },
    {
      "epoch": 0.00945,
      "grad_norm": 0.760430287307007,
      "learning_rate": 0.002835,
      "loss": 4.567,
      "step": 945
    },
    {
      "epoch": 0.00946,
      "grad_norm": 0.8410168172764453,
      "learning_rate": 0.002838,
      "loss": 4.5808,
      "step": 946
    },
    {
      "epoch": 0.00947,
      "grad_norm": 0.8502364092306604,
      "learning_rate": 0.0028409999999999998,
      "loss": 4.5581,
      "step": 947
    },
    {
      "epoch": 0.00948,
      "grad_norm": 0.7534324730199542,
      "learning_rate": 0.0028439999999999997,
      "loss": 4.533,
      "step": 948
    },
    {
      "epoch": 0.00949,
      "grad_norm": 0.8075715283027973,
      "learning_rate": 0.002847,
      "loss": 4.5789,
      "step": 949
    },
    {
      "epoch": 0.0095,
      "grad_norm": 0.8790685187514339,
      "learning_rate": 0.00285,
      "loss": 4.5764,
      "step": 950
    },
    {
      "epoch": 0.00951,
      "grad_norm": 0.8527621336415785,
      "learning_rate": 0.002853,
      "loss": 4.552,
      "step": 951
    },
    {
      "epoch": 0.00952,
      "grad_norm": 0.793648162843131,
      "learning_rate": 0.002856,
      "loss": 4.5771,
      "step": 952
    },
    {
      "epoch": 0.00953,
      "grad_norm": 0.7100823051409002,
      "learning_rate": 0.002859,
      "loss": 4.5151,
      "step": 953
    },
    {
      "epoch": 0.00954,
      "grad_norm": 0.776086010454581,
      "learning_rate": 0.002862,
      "loss": 4.5748,
      "step": 954
    },
    {
      "epoch": 0.00955,
      "grad_norm": 0.7357834745016256,
      "learning_rate": 0.002865,
      "loss": 4.5651,
      "step": 955
    },
    {
      "epoch": 0.00956,
      "grad_norm": 0.6871788604084053,
      "learning_rate": 0.002868,
      "loss": 4.5378,
      "step": 956
    },
    {
      "epoch": 0.00957,
      "grad_norm": 0.6293704920093642,
      "learning_rate": 0.002871,
      "loss": 4.5585,
      "step": 957
    },
    {
      "epoch": 0.00958,
      "grad_norm": 0.6933721545151298,
      "learning_rate": 0.002874,
      "loss": 4.5402,
      "step": 958
    },
    {
      "epoch": 0.00959,
      "grad_norm": 0.6216290945191316,
      "learning_rate": 0.002877,
      "loss": 4.5294,
      "step": 959
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.44090482568449035,
      "learning_rate": 0.0028799999999999997,
      "loss": 4.5205,
      "step": 960
    },
    {
      "epoch": 0.00961,
      "grad_norm": 0.5026549244936088,
      "learning_rate": 0.002883,
      "loss": 4.4973,
      "step": 961
    },
    {
      "epoch": 0.00962,
      "grad_norm": 0.46550744372429714,
      "learning_rate": 0.002886,
      "loss": 4.5199,
      "step": 962
    },
    {
      "epoch": 0.00963,
      "grad_norm": 0.4817462883709995,
      "learning_rate": 0.002889,
      "loss": 4.5204,
      "step": 963
    },
    {
      "epoch": 0.00964,
      "grad_norm": 0.5021989893794454,
      "learning_rate": 0.002892,
      "loss": 4.5105,
      "step": 964
    },
    {
      "epoch": 0.00965,
      "grad_norm": 0.6331237702649058,
      "learning_rate": 0.002895,
      "loss": 4.4888,
      "step": 965
    },
    {
      "epoch": 0.00966,
      "grad_norm": 0.7186463225121739,
      "learning_rate": 0.002898,
      "loss": 4.5122,
      "step": 966
    },
    {
      "epoch": 0.00967,
      "grad_norm": 0.835541291398658,
      "learning_rate": 0.002901,
      "loss": 4.5497,
      "step": 967
    },
    {
      "epoch": 0.00968,
      "grad_norm": 0.7770950591910699,
      "learning_rate": 0.002904,
      "loss": 4.5246,
      "step": 968
    },
    {
      "epoch": 0.00969,
      "grad_norm": 0.6421972738290654,
      "learning_rate": 0.002907,
      "loss": 4.5465,
      "step": 969
    },
    {
      "epoch": 0.0097,
      "grad_norm": 0.6170493579190435,
      "learning_rate": 0.00291,
      "loss": 4.5014,
      "step": 970
    },
    {
      "epoch": 0.00971,
      "grad_norm": 0.671893763395282,
      "learning_rate": 0.002913,
      "loss": 4.5134,
      "step": 971
    },
    {
      "epoch": 0.00972,
      "grad_norm": 0.5384349268117217,
      "learning_rate": 0.002916,
      "loss": 4.51,
      "step": 972
    },
    {
      "epoch": 0.00973,
      "grad_norm": 0.6570052261370841,
      "learning_rate": 0.002919,
      "loss": 4.5075,
      "step": 973
    },
    {
      "epoch": 0.00974,
      "grad_norm": 0.6469437996214488,
      "learning_rate": 0.002922,
      "loss": 4.5042,
      "step": 974
    },
    {
      "epoch": 0.00975,
      "grad_norm": 0.5139434995269291,
      "learning_rate": 0.002925,
      "loss": 4.5141,
      "step": 975
    },
    {
      "epoch": 0.00976,
      "grad_norm": 0.540350404123188,
      "learning_rate": 0.002928,
      "loss": 4.4984,
      "step": 976
    },
    {
      "epoch": 0.00977,
      "grad_norm": 0.5640158884340003,
      "learning_rate": 0.002931,
      "loss": 4.5359,
      "step": 977
    },
    {
      "epoch": 0.00978,
      "grad_norm": 0.5471232322596488,
      "learning_rate": 0.002934,
      "loss": 4.5069,
      "step": 978
    },
    {
      "epoch": 0.00979,
      "grad_norm": 0.6391692840311302,
      "learning_rate": 0.002937,
      "loss": 4.4972,
      "step": 979
    },
    {
      "epoch": 0.0098,
      "grad_norm": 0.7065424241899814,
      "learning_rate": 0.00294,
      "loss": 4.4963,
      "step": 980
    },
    {
      "epoch": 0.00981,
      "grad_norm": 0.6903724510426201,
      "learning_rate": 0.002943,
      "loss": 4.5078,
      "step": 981
    },
    {
      "epoch": 0.00982,
      "grad_norm": 0.653302049548968,
      "learning_rate": 0.002946,
      "loss": 4.4991,
      "step": 982
    },
    {
      "epoch": 0.00983,
      "grad_norm": 0.7786836590229197,
      "learning_rate": 0.0029490000000000002,
      "loss": 4.5061,
      "step": 983
    },
    {
      "epoch": 0.00984,
      "grad_norm": 0.8944946184941699,
      "learning_rate": 0.002952,
      "loss": 4.5043,
      "step": 984
    },
    {
      "epoch": 0.00985,
      "grad_norm": 0.9434362337974719,
      "learning_rate": 0.002955,
      "loss": 4.5335,
      "step": 985
    },
    {
      "epoch": 0.00986,
      "grad_norm": 0.9312010498282314,
      "learning_rate": 0.002958,
      "loss": 4.4908,
      "step": 986
    },
    {
      "epoch": 0.00987,
      "grad_norm": 0.7600059379317153,
      "learning_rate": 0.002961,
      "loss": 4.5038,
      "step": 987
    },
    {
      "epoch": 0.00988,
      "grad_norm": 0.6949983731085813,
      "learning_rate": 0.002964,
      "loss": 4.533,
      "step": 988
    },
    {
      "epoch": 0.00989,
      "grad_norm": 0.6958207764457225,
      "learning_rate": 0.002967,
      "loss": 4.4744,
      "step": 989
    },
    {
      "epoch": 0.0099,
      "grad_norm": 0.8314159874496235,
      "learning_rate": 0.00297,
      "loss": 4.5218,
      "step": 990
    },
    {
      "epoch": 0.00991,
      "grad_norm": 0.90109982550625,
      "learning_rate": 0.002973,
      "loss": 4.5224,
      "step": 991
    },
    {
      "epoch": 0.00992,
      "grad_norm": 0.951208720196979,
      "learning_rate": 0.002976,
      "loss": 4.5281,
      "step": 992
    },
    {
      "epoch": 0.00993,
      "grad_norm": 1.034654706401041,
      "learning_rate": 0.002979,
      "loss": 4.539,
      "step": 993
    },
    {
      "epoch": 0.00994,
      "grad_norm": 1.0078120294806783,
      "learning_rate": 0.002982,
      "loss": 4.5459,
      "step": 994
    },
    {
      "epoch": 0.00995,
      "grad_norm": 0.9905056664043064,
      "learning_rate": 0.0029850000000000002,
      "loss": 4.5461,
      "step": 995
    },
    {
      "epoch": 0.00996,
      "grad_norm": 1.1603429750232241,
      "learning_rate": 0.002988,
      "loss": 4.5605,
      "step": 996
    },
    {
      "epoch": 0.00997,
      "grad_norm": 0.8595677567729485,
      "learning_rate": 0.002991,
      "loss": 4.5366,
      "step": 997
    },
    {
      "epoch": 0.00998,
      "grad_norm": 0.9672785170741734,
      "learning_rate": 0.002994,
      "loss": 4.5467,
      "step": 998
    },
    {
      "epoch": 0.00999,
      "grad_norm": 0.8855476122040005,
      "learning_rate": 0.002997,
      "loss": 4.5306,
      "step": 999
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7479993027173912,
      "learning_rate": 0.003,
      "loss": 4.5436,
      "step": 1000
    }
  ],
  "logging_steps": 1,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.9643642855424e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}