{
  "best_global_step": 4952,
  "best_metric": 2.4348788261413574,
  "best_model_checkpoint": "./output_dir/Video_Games_stage1_full_rollthenthink/checkpoint-4952",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 4952,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00040401999898995,
      "grad_norm": 34.5,
      "learning_rate": 0.0,
      "loss": 2.5641,
      "step": 1
    },
    {
      "epoch": 0.0008080399979799,
      "grad_norm": 25.625,
      "learning_rate": 1.4999999999999999e-05,
      "loss": 1.8995,
      "step": 2
    },
    {
      "epoch": 0.00121205999696985,
      "grad_norm": 14.75,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 1.6299,
      "step": 3
    },
    {
      "epoch": 0.0016160799959598,
      "grad_norm": 8.9375,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 1.8235,
      "step": 4
    },
    {
      "epoch": 0.0020200999949497502,
      "grad_norm": 3.140625,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 1.4428,
      "step": 5
    },
    {
      "epoch": 0.0024241199939397,
      "grad_norm": 3.359375,
      "learning_rate": 7.5e-05,
      "loss": 1.6732,
      "step": 6
    },
    {
      "epoch": 0.00282813999292965,
      "grad_norm": 2.578125,
      "learning_rate": 8.999999999999999e-05,
      "loss": 1.4055,
      "step": 7
    },
    {
      "epoch": 0.0032321599919196,
      "grad_norm": 1.765625,
      "learning_rate": 0.00010499999999999999,
      "loss": 1.3038,
      "step": 8
    },
    {
      "epoch": 0.00363617999090955,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00011999999999999999,
      "loss": 1.4008,
      "step": 9
    },
    {
      "epoch": 0.0040401999898995004,
      "grad_norm": 1.9140625,
      "learning_rate": 0.000135,
      "loss": 1.3479,
      "step": 10
    },
    {
      "epoch": 0.00444421998888945,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00015,
      "loss": 1.0933,
      "step": 11
    },
    {
      "epoch": 0.0048482399878794,
      "grad_norm": 1.59375,
      "learning_rate": 0.000165,
      "loss": 1.2659,
      "step": 12
    },
    {
      "epoch": 0.00525225998686935,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00017999999999999998,
      "loss": 1.2401,
      "step": 13
    },
    {
      "epoch": 0.0056562799858593,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000195,
      "loss": 1.2296,
      "step": 14
    },
    {
      "epoch": 0.00606029998484925,
      "grad_norm": 1.203125,
      "learning_rate": 0.00020999999999999998,
      "loss": 1.0735,
      "step": 15
    },
    {
      "epoch": 0.0064643199838392,
      "grad_norm": 1.3125,
      "learning_rate": 0.000225,
      "loss": 1.2826,
      "step": 16
    },
    {
      "epoch": 0.00686833998282915,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00023999999999999998,
      "loss": 1.0531,
      "step": 17
    },
    {
      "epoch": 0.0072723599818191,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025499999999999996,
      "loss": 1.2165,
      "step": 18
    },
    {
      "epoch": 0.00767637998080905,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027,
      "loss": 1.1774,
      "step": 19
    },
    {
      "epoch": 0.008080399979799001,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000285,
      "loss": 1.1318,
      "step": 20
    },
    {
      "epoch": 0.008484419978788951,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003,
      "loss": 1.0383,
      "step": 21
    },
    {
      "epoch": 0.0088884399777789,
      "grad_norm": 1.4375,
      "learning_rate": 0.00029997572815533975,
      "loss": 1.0022,
      "step": 22
    },
    {
      "epoch": 0.00929245997676885,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002999514563106796,
      "loss": 1.2895,
      "step": 23
    },
    {
      "epoch": 0.0096964799757588,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002999271844660194,
      "loss": 1.0691,
      "step": 24
    },
    {
      "epoch": 0.01010049997474875,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002999029126213592,
      "loss": 1.2265,
      "step": 25
    },
    {
      "epoch": 0.0105045199737387,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000299878640776699,
      "loss": 1.288,
      "step": 26
    },
    {
      "epoch": 0.01090853997272865,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002998543689320388,
      "loss": 1.0984,
      "step": 27
    },
    {
      "epoch": 0.0113125599717186,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002998300970873786,
      "loss": 1.3475,
      "step": 28
    },
    {
      "epoch": 0.01171657997070855,
      "grad_norm": 1.453125,
      "learning_rate": 0.00029980582524271845,
      "loss": 1.3382,
      "step": 29
    },
    {
      "epoch": 0.0121205999696985,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002997815533980582,
      "loss": 1.3884,
      "step": 30
    },
    {
      "epoch": 0.01252461996868845,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00029975728155339805,
      "loss": 1.0333,
      "step": 31
    },
    {
      "epoch": 0.0129286399676784,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002997330097087378,
      "loss": 1.1701,
      "step": 32
    },
    {
      "epoch": 0.01333265996666835,
      "grad_norm": 3.65625,
      "learning_rate": 0.00029970873786407766,
      "loss": 1.1252,
      "step": 33
    },
    {
      "epoch": 0.0137366799656583,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029968446601941743,
      "loss": 1.1997,
      "step": 34
    },
    {
      "epoch": 0.01414069996464825,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00029966019417475726,
      "loss": 1.183,
      "step": 35
    },
    {
      "epoch": 0.0145447199636382,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029963592233009704,
      "loss": 1.1745,
      "step": 36
    },
    {
      "epoch": 0.01494873996262815,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00029961165048543687,
      "loss": 1.2271,
      "step": 37
    },
    {
      "epoch": 0.0153527599616181,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029958737864077664,
      "loss": 1.1588,
      "step": 38
    },
    {
      "epoch": 0.01575677996060805,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029956310679611647,
      "loss": 1.2681,
      "step": 39
    },
    {
      "epoch": 0.016160799959598002,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002995388349514563,
      "loss": 1.0733,
      "step": 40
    },
    {
      "epoch": 0.01656481995858795,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002995145631067961,
      "loss": 0.9785,
      "step": 41
    },
    {
      "epoch": 0.016968839957577902,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002994902912621359,
      "loss": 1.1494,
      "step": 42
    },
    {
      "epoch": 0.01737285995656785,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002994660194174757,
      "loss": 1.1549,
      "step": 43
    },
    {
      "epoch": 0.0177768799555578,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002994417475728155,
      "loss": 1.2268,
      "step": 44
    },
    {
      "epoch": 0.01818089995454775,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029941747572815534,
      "loss": 1.1043,
      "step": 45
    },
    {
      "epoch": 0.0185849199535377,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002993932038834951,
      "loss": 1.3574,
      "step": 46
    },
    {
      "epoch": 0.01898893995252765,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029936893203883494,
      "loss": 1.2589,
      "step": 47
    },
    {
      "epoch": 0.0193929599515176,
      "grad_norm": 1.0,
      "learning_rate": 0.0002993446601941747,
      "loss": 1.0932,
      "step": 48
    },
    {
      "epoch": 0.01979697995050755,
      "grad_norm": 0.90625,
      "learning_rate": 0.00029932038834951455,
      "loss": 1.053,
      "step": 49
    },
    {
      "epoch": 0.0202009999494975,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002992961165048544,
      "loss": 1.0963,
      "step": 50
    },
    {
      "epoch": 0.02060501994848745,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029927184466019415,
      "loss": 1.2248,
      "step": 51
    },
    {
      "epoch": 0.0210090399474774,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029924757281553393,
      "loss": 1.1382,
      "step": 52
    },
    {
      "epoch": 0.02141305994646735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029922330097087376,
      "loss": 1.2549,
      "step": 53
    },
    {
      "epoch": 0.0218170799454573,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002991990291262136,
      "loss": 1.2379,
      "step": 54
    },
    {
      "epoch": 0.02222109994444725,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029917475728155336,
      "loss": 1.2471,
      "step": 55
    },
    {
      "epoch": 0.0226251199434372,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002991504854368932,
      "loss": 1.3349,
      "step": 56
    },
    {
      "epoch": 0.02302913994242715,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029912621359223297,
      "loss": 1.2337,
      "step": 57
    },
    {
      "epoch": 0.0234331599414171,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002991019417475728,
      "loss": 1.297,
      "step": 58
    },
    {
      "epoch": 0.02383717994040705,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029907766990291257,
      "loss": 1.1379,
      "step": 59
    },
    {
      "epoch": 0.024241199939397,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002990533980582524,
      "loss": 1.0976,
      "step": 60
    },
    {
      "epoch": 0.02464521993838695,
      "grad_norm": 1.15625,
      "learning_rate": 0.00029902912621359223,
      "loss": 1.2591,
      "step": 61
    },
    {
      "epoch": 0.0250492399373769,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000299004854368932,
      "loss": 1.1093,
      "step": 62
    },
    {
      "epoch": 0.02545325993636685,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002989805825242718,
      "loss": 1.3025,
      "step": 63
    },
    {
      "epoch": 0.0258572799353568,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002989563106796116,
      "loss": 0.9812,
      "step": 64
    },
    {
      "epoch": 0.02626129993434675,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029893203883495144,
      "loss": 1.1996,
      "step": 65
    },
    {
      "epoch": 0.0266653199333367,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002989077669902912,
      "loss": 1.2251,
      "step": 66
    },
    {
      "epoch": 0.02706933993232665,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00029888349514563104,
      "loss": 1.002,
      "step": 67
    },
    {
      "epoch": 0.0274733599313166,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002988592233009708,
      "loss": 1.1654,
      "step": 68
    },
    {
      "epoch": 0.02787737993030655,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029883495145631065,
      "loss": 1.308,
      "step": 69
    },
    {
      "epoch": 0.0282813999292965,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002988106796116505,
      "loss": 1.3042,
      "step": 70
    },
    {
      "epoch": 0.02868541992828645,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029878640776699025,
      "loss": 1.1926,
      "step": 71
    },
    {
      "epoch": 0.0290894399272764,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002987621359223301,
      "loss": 1.1621,
      "step": 72
    },
    {
      "epoch": 0.02949345992626635,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00029873786407766986,
      "loss": 1.1024,
      "step": 73
    },
    {
      "epoch": 0.0298974799252563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002987135922330097,
      "loss": 1.2151,
      "step": 74
    },
    {
      "epoch": 0.03030149992424625,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002986893203883495,
      "loss": 1.0686,
      "step": 75
    },
    {
      "epoch": 0.0307055199232362,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002986650485436893,
      "loss": 1.0877,
      "step": 76
    },
    {
      "epoch": 0.03110953992222615,
      "grad_norm": 1.0,
      "learning_rate": 0.0002986407766990291,
      "loss": 1.2617,
      "step": 77
    },
    {
      "epoch": 0.0315135599212161,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002986165048543689,
      "loss": 0.9851,
      "step": 78
    },
    {
      "epoch": 0.03191757992020605,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002985922330097087,
      "loss": 1.031,
      "step": 79
    },
    {
      "epoch": 0.032321599919196004,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029856796116504856,
      "loss": 1.323,
      "step": 80
    },
    {
      "epoch": 0.03272561991818595,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029854368932038833,
      "loss": 1.2017,
      "step": 81
    },
    {
      "epoch": 0.0331296399171759,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002985194174757281,
      "loss": 1.1801,
      "step": 82
    },
    {
      "epoch": 0.03353365991616585,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029849514563106794,
      "loss": 1.0822,
      "step": 83
    },
    {
      "epoch": 0.033937679915155804,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029847087378640777,
      "loss": 1.1898,
      "step": 84
    },
    {
      "epoch": 0.03434169991414575,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029844660194174754,
      "loss": 1.0311,
      "step": 85
    },
    {
      "epoch": 0.0347457199131357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029842233009708737,
      "loss": 1.2454,
      "step": 86
    },
    {
      "epoch": 0.03514973991212565,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00029839805825242715,
      "loss": 1.0596,
      "step": 87
    },
    {
      "epoch": 0.0355537599111156,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000298373786407767,
      "loss": 1.0798,
      "step": 88
    },
    {
      "epoch": 0.03595777991010555,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029834951456310675,
      "loss": 1.172,
      "step": 89
    },
    {
      "epoch": 0.0363617999090955,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002983252427184466,
      "loss": 1.1422,
      "step": 90
    },
    {
      "epoch": 0.03676581990808545,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002983009708737864,
      "loss": 1.1116,
      "step": 91
    },
    {
      "epoch": 0.0371698399070754,
      "grad_norm": 0.875,
      "learning_rate": 0.0002982766990291262,
      "loss": 1.0879,
      "step": 92
    },
    {
      "epoch": 0.03757385990606535,
      "grad_norm": 0.875,
      "learning_rate": 0.00029825242718446596,
      "loss": 0.9441,
      "step": 93
    },
    {
      "epoch": 0.0379778799050553,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0002982281553398058,
      "loss": 0.9607,
      "step": 94
    },
    {
      "epoch": 0.03838189990404525,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0002982038834951456,
      "loss": 1.0844,
      "step": 95
    },
    {
      "epoch": 0.0387859199030352,
      "grad_norm": 1.0,
      "learning_rate": 0.0002981796116504854,
      "loss": 1.0862,
      "step": 96
    },
    {
      "epoch": 0.03918993990202515,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0002981553398058252,
      "loss": 1.0541,
      "step": 97
    },
    {
      "epoch": 0.0395939599010151,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000298131067961165,
      "loss": 1.1086,
      "step": 98
    },
    {
      "epoch": 0.03999797990000505,
      "grad_norm": 0.859375,
      "learning_rate": 0.00029810679611650483,
      "loss": 1.034,
      "step": 99
    },
    {
      "epoch": 0.040401999898995,
      "grad_norm": 0.8125,
      "learning_rate": 0.00029808252427184466,
      "loss": 0.9623,
      "step": 100
    },
    {
      "epoch": 0.04080601989798495,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00029805825242718443,
      "loss": 1.1094,
      "step": 101
    },
    {
      "epoch": 0.0412100398969749,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00029803398058252426,
      "loss": 0.9687,
      "step": 102
    },
    {
      "epoch": 0.04161405989596485,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00029800970873786404,
      "loss": 0.9616,
      "step": 103
    },
    {
      "epoch": 0.0420180798949548,
      "grad_norm": 0.8125,
      "learning_rate": 0.00029798543689320387,
      "loss": 1.0682,
      "step": 104
    },
    {
      "epoch": 0.04242209989394475,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002979611650485437,
      "loss": 0.8879,
      "step": 105
    },
    {
      "epoch": 0.0428261198929347,
      "grad_norm": 0.796875,
      "learning_rate": 0.00029793689320388347,
      "loss": 1.0175,
      "step": 106
    },
    {
      "epoch": 0.04323013989192465,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002979126213592233,
      "loss": 1.1171,
      "step": 107
    },
    {
      "epoch": 0.0436341598909146,
      "grad_norm": 0.765625,
      "learning_rate": 0.0002978883495145631,
      "loss": 0.9731,
      "step": 108
    },
    {
      "epoch": 0.04403817988990455,
      "grad_norm": 0.8125,
      "learning_rate": 0.0002978640776699029,
      "loss": 1.0341,
      "step": 109
    },
    {
      "epoch": 0.0444421998888945,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002978398058252427,
      "loss": 1.0358,
      "step": 110
    },
    {
      "epoch": 0.04484621988788445,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0002978155339805825,
      "loss": 1.0316,
      "step": 111
    },
    {
      "epoch": 0.0452502398868744,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002977912621359223,
      "loss": 1.0325,
      "step": 112
    },
    {
      "epoch": 0.04565425988586435,
      "grad_norm": 0.890625,
      "learning_rate": 0.0002977669902912621,
      "loss": 1.1484,
      "step": 113
    },
    {
      "epoch": 0.0460582798848543,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002977427184466019,
      "loss": 0.9972,
      "step": 114
    },
    {
      "epoch": 0.04646229988384425,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0002977184466019417,
      "loss": 0.9471,
      "step": 115
    },
    {
      "epoch": 0.0468663198828342,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00029769417475728155,
      "loss": 0.9588,
      "step": 116
    },
    {
      "epoch": 0.04727033988182415,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0002976699029126213,
      "loss": 1.1364,
      "step": 117
    },
    {
      "epoch": 0.0476743598808141,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00029764563106796115,
      "loss": 0.9852,
      "step": 118
    },
    {
      "epoch": 0.048078379879804053,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00029762135922330093,
      "loss": 1.1316,
      "step": 119
    },
    {
      "epoch": 0.048482399878794,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029759708737864076,
      "loss": 1.1558,
      "step": 120
    },
    {
      "epoch": 0.04888641987778395,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0002975728155339806,
      "loss": 1.0148,
      "step": 121
    },
    {
      "epoch": 0.0492904398767739,
      "grad_norm": 0.859375,
      "learning_rate": 0.00029754854368932036,
      "loss": 1.0293,
      "step": 122
    },
    {
      "epoch": 0.049694459875763854,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00029752427184466014,
      "loss": 1.0008,
      "step": 123
    },
    {
      "epoch": 0.0500984798747538,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00029749999999999997,
      "loss": 0.9754,
      "step": 124
    },
    {
      "epoch": 0.05050249987374375,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002974757281553398,
      "loss": 1.0863,
      "step": 125
    },
    {
      "epoch": 0.0509065198727337,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00029745145631067957,
      "loss": 1.0305,
      "step": 126
    },
    {
      "epoch": 0.05131053987172365,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002974271844660194,
      "loss": 0.9882,
      "step": 127
    },
    {
      "epoch": 0.0517145598707136,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002974029126213592,
      "loss": 1.1303,
      "step": 128
    },
    {
      "epoch": 0.05211857986970355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000297378640776699,
      "loss": 1.1039,
      "step": 129
    },
    {
      "epoch": 0.0525225998686935,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00029735436893203884,
      "loss": 1.1055,
      "step": 130
    },
    {
      "epoch": 0.05292661986768345,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002973300970873786,
      "loss": 1.133,
      "step": 131
    },
    {
      "epoch": 0.0533306398666734,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00029730582524271844,
      "loss": 1.0825,
      "step": 132
    },
    {
      "epoch": 0.05373465986566335,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0002972815533980582,
      "loss": 0.9782,
      "step": 133
    },
    {
      "epoch": 0.0541386798646533,
      "grad_norm": 0.875,
      "learning_rate": 0.00029725728155339804,
      "loss": 0.9056,
      "step": 134
    },
    {
      "epoch": 0.05454269986364325,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002972330097087378,
      "loss": 1.0426,
      "step": 135
    },
    {
      "epoch": 0.0549467198626332,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029720873786407765,
      "loss": 1.1402,
      "step": 136
    },
    {
      "epoch": 0.05535073986162315,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0002971844660194174,
      "loss": 0.9809,
      "step": 137
    },
    {
      "epoch": 0.0557547598606131,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00029716019417475725,
      "loss": 0.9833,
      "step": 138
    },
    {
      "epoch": 0.05615877985960305,
      "grad_norm": 0.859375,
      "learning_rate": 0.00029713592233009703,
      "loss": 1.0878,
      "step": 139
    },
    {
      "epoch": 0.056562799858593,
      "grad_norm": 0.828125,
      "learning_rate": 0.00029711165048543686,
      "loss": 1.0767,
      "step": 140
    },
    {
      "epoch": 0.05696681985758295,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0002970873786407767,
      "loss": 0.9885,
      "step": 141
    },
    {
      "epoch": 0.0573708398565729,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00029706310679611646,
      "loss": 0.8937,
      "step": 142
    },
    {
      "epoch": 0.05777485985556285,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002970388349514563,
      "loss": 1.1756,
      "step": 143
    },
    {
      "epoch": 0.0581788798545528,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00029701456310679607,
      "loss": 0.9337,
      "step": 144
    },
    {
      "epoch": 0.05858289985354275,
      "grad_norm": 0.890625,
      "learning_rate": 0.0002969902912621359,
      "loss": 1.033,
      "step": 145
    },
    {
      "epoch": 0.0589869198525327,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002969660194174757,
      "loss": 1.2059,
      "step": 146
    },
    {
      "epoch": 0.05939093985152265,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002969417475728155,
      "loss": 1.0209,
      "step": 147
    },
    {
      "epoch": 0.0597949598505126,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00029691747572815533,
      "loss": 1.0435,
      "step": 148
    },
    {
      "epoch": 0.06019897984950255,
      "grad_norm": 0.875,
      "learning_rate": 0.0002968932038834951,
      "loss": 0.973,
      "step": 149
    },
    {
      "epoch": 0.0606029998484925,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00029686893203883494,
      "loss": 0.9803,
      "step": 150
    },
    {
      "epoch": 0.06100701984748245,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00029684466019417477,
      "loss": 0.9996,
      "step": 151
    },
    {
      "epoch": 0.0614110398464724,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00029682038834951454,
      "loss": 1.0078,
      "step": 152
    },
    {
      "epoch": 0.06181505984546235,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0002967961165048543,
      "loss": 1.0159,
      "step": 153
    },
    {
      "epoch": 0.0622190798444523,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00029677184466019415,
      "loss": 0.8266,
      "step": 154
    },
    {
      "epoch": 0.06262309984344225,
      "grad_norm": 0.84375,
      "learning_rate": 0.000296747572815534,
      "loss": 0.9275,
      "step": 155
    },
    {
      "epoch": 0.0630271198424322,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029672330097087375,
      "loss": 1.0159,
      "step": 156
    },
    {
      "epoch": 0.06343113984142215,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002966990291262136,
      "loss": 1.1252,
      "step": 157
    },
    {
      "epoch": 0.0638351598404121,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029667475728155336,
      "loss": 1.1385,
      "step": 158
    },
    {
      "epoch": 0.06423917983940206,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0002966504854368932,
      "loss": 1.032,
      "step": 159
    },
    {
      "epoch": 0.06464319983839201,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00029662621359223296,
      "loss": 0.9425,
      "step": 160
    },
    {
      "epoch": 0.06504721983738194,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002966019417475728,
      "loss": 1.0501,
      "step": 161
    },
    {
      "epoch": 0.0654512398363719,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0002965776699029126,
      "loss": 0.9881,
      "step": 162
    },
    {
      "epoch": 0.06585525983536185,
      "grad_norm": 0.796875,
      "learning_rate": 0.0002965533980582524,
      "loss": 0.931,
      "step": 163
    },
    {
      "epoch": 0.0662592798343518,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00029652912621359217,
      "loss": 0.933,
      "step": 164
    },
    {
      "epoch": 0.06666329983334175,
      "grad_norm": 0.7109375,
      "learning_rate": 0.000296504854368932,
      "loss": 0.9364,
      "step": 165
    },
    {
      "epoch": 0.0670673198323317,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00029648058252427183,
      "loss": 1.0804,
      "step": 166
    },
    {
      "epoch": 0.06747133983132166,
      "grad_norm": 0.78125,
      "learning_rate": 0.0002964563106796116,
      "loss": 0.9827,
      "step": 167
    },
    {
      "epoch": 0.06787535983031161,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029643203883495143,
      "loss": 1.1814,
      "step": 168
    },
    {
      "epoch": 0.06827937982930155,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002964077669902912,
      "loss": 0.9523,
      "step": 169
    },
    {
      "epoch": 0.0686833998282915,
      "grad_norm": 0.76953125,
      "learning_rate": 0.00029638349514563104,
      "loss": 0.9115,
      "step": 170
    },
    {
      "epoch": 0.06908741982728145,
      "grad_norm": 0.73828125,
      "learning_rate": 0.00029635922330097087,
      "loss": 0.9798,
      "step": 171
    },
    {
      "epoch": 0.0694914398262714,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00029633495145631064,
      "loss": 1.0933,
      "step": 172
    },
    {
      "epoch": 0.06989545982526135,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00029631067961165047,
      "loss": 1.0964,
      "step": 173
    },
    {
      "epoch": 0.0702994798242513,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00029628640776699025,
      "loss": 0.8969,
      "step": 174
    },
    {
      "epoch": 0.07070349982324126,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002962621359223301,
      "loss": 1.0792,
      "step": 175
    },
    {
      "epoch": 0.0711075198222312,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0002962378640776699,
      "loss": 1.1553,
      "step": 176
    },
    {
      "epoch": 0.07151153982122115,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002962135922330097,
      "loss": 0.851,
      "step": 177
    },
    {
      "epoch": 0.0719155598202111,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0002961893203883495,
      "loss": 0.8923,
      "step": 178
    },
    {
      "epoch": 0.07231957981920105,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0002961650485436893,
      "loss": 1.0365,
      "step": 179
    },
    {
      "epoch": 0.072723599818191,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0002961407766990291,
      "loss": 0.9531,
      "step": 180
    },
    {
      "epoch": 0.07312761981718095,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00029611650485436894,
      "loss": 1.011,
      "step": 181
    },
    {
      "epoch": 0.0735316398161709,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0002960922330097087,
      "loss": 1.0127,
      "step": 182
    },
    {
      "epoch": 0.07393565981516086,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0002960679611650485,
      "loss": 1.092,
      "step": 183
    },
    {
      "epoch": 0.0743396798141508,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0002960436893203883,
      "loss": 1.0027,
      "step": 184
    },
    {
      "epoch": 0.07474369981314075,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002960194174757281,
      "loss": 0.8608,
      "step": 185
    },
    {
      "epoch": 0.0751477198121307,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00029599514563106793,
      "loss": 1.1078,
      "step": 186
    },
    {
      "epoch": 0.07555173981112065,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00029597087378640776,
      "loss": 1.1133,
      "step": 187
    },
    {
      "epoch": 0.0759557598101106,
      "grad_norm": 0.90625,
      "learning_rate": 0.00029594660194174753,
      "loss": 0.9431,
      "step": 188
    },
    {
      "epoch": 0.07635977980910055,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00029592233009708736,
      "loss": 1.0513,
      "step": 189
    },
    {
      "epoch": 0.0767637998080905,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00029589805825242714,
      "loss": 1.1064,
      "step": 190
    },
    {
      "epoch": 0.07716781980708046,
      "grad_norm": 0.6875,
      "learning_rate": 0.00029587378640776697,
      "loss": 0.889,
      "step": 191
    },
    {
      "epoch": 0.0775718398060704,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0002958495145631068,
      "loss": 0.832,
      "step": 192
    },
    {
      "epoch": 0.07797585980506035,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00029582524271844657,
      "loss": 0.8738,
      "step": 193
    },
    {
      "epoch": 0.0783798798040503,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00029580097087378635,
      "loss": 1.1953,
      "step": 194
    },
    {
      "epoch": 0.07878389980304025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002957766990291262,
      "loss": 1.2061,
      "step": 195
    },
    {
      "epoch": 0.0791879198020302,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000295752427184466,
      "loss": 1.1278,
      "step": 196
    },
    {
      "epoch": 0.07959193980102015,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0002957281553398058,
      "loss": 0.9646,
      "step": 197
    },
    {
      "epoch": 0.0799959598000101,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0002957038834951456,
      "loss": 1.0379,
      "step": 198
    },
    {
      "epoch": 0.08039997979900006,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0002956796116504854,
      "loss": 1.1067,
      "step": 199
    },
    {
      "epoch": 0.08080399979799,
      "grad_norm": 0.75,
      "learning_rate": 0.0002956553398058252,
      "loss": 0.9496,
      "step": 200
    },
    {
      "epoch": 0.08120801979697995,
      "grad_norm": 0.71875,
      "learning_rate": 0.00029563106796116505,
      "loss": 0.9241,
      "step": 201
    },
    {
      "epoch": 0.0816120397959699,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002956067961165048,
      "loss": 0.9225,
      "step": 202
    },
    {
      "epoch": 0.08201605979495985,
      "grad_norm": 0.828125,
      "learning_rate": 0.00029558252427184465,
      "loss": 1.1124,
      "step": 203
    },
    {
      "epoch": 0.0824200797939498,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0002955582524271844,
      "loss": 1.0216,
      "step": 204
    },
    {
      "epoch": 0.08282409979293975,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00029553398058252425,
      "loss": 1.1123,
      "step": 205
    },
    {
      "epoch": 0.0832281197919297,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0002955097087378641,
      "loss": 0.9313,
      "step": 206
    },
    {
      "epoch": 0.08363213979091966,
      "grad_norm": 0.90625,
      "learning_rate": 0.00029548543689320386,
      "loss": 0.9868,
      "step": 207
    },
    {
      "epoch": 0.0840361597899096,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0002954611650485437,
      "loss": 0.8957,
      "step": 208
    },
    {
      "epoch": 0.08444017978889955,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00029543689320388346,
      "loss": 1.0486,
      "step": 209
    },
    {
      "epoch": 0.0848441997878895,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002954126213592233,
      "loss": 0.9594,
      "step": 210
    },
    {
      "epoch": 0.08524821978687945,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00029538834951456307,
      "loss": 1.0598,
      "step": 211
    },
    {
      "epoch": 0.0856522397858694,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002953640776699029,
      "loss": 0.9287,
      "step": 212
    },
    {
      "epoch": 0.08605625978485935,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0002953398058252427,
      "loss": 0.9725,
      "step": 213
    },
    {
      "epoch": 0.0864602797838493,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0002953155339805825,
      "loss": 0.9781,
      "step": 214
    },
    {
      "epoch": 0.08686429978283924,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0002952912621359223,
      "loss": 0.8706,
      "step": 215
    },
    {
      "epoch": 0.0872683197818292,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0002952669902912621,
      "loss": 0.9089,
      "step": 216
    },
    {
      "epoch": 0.08767233978081915,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00029524271844660194,
      "loss": 1.0379,
      "step": 217
    },
    {
      "epoch": 0.0880763597798091,
      "grad_norm": 0.765625,
      "learning_rate": 0.0002952184466019417,
      "loss": 1.0308,
      "step": 218
    },
    {
      "epoch": 0.08848037977879905,
      "grad_norm": 0.7421875,
      "learning_rate": 0.00029519417475728154,
      "loss": 0.9835,
      "step": 219
    },
    {
      "epoch": 0.088884399777789,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002951699029126213,
      "loss": 0.9932,
      "step": 220
    },
    {
      "epoch": 0.08928841977677895,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00029514563106796115,
      "loss": 1.0426,
      "step": 221
    },
    {
      "epoch": 0.0896924397757689,
      "grad_norm": 0.74609375,
      "learning_rate": 0.000295121359223301,
      "loss": 0.9382,
      "step": 222
    },
    {
      "epoch": 0.09009645977475884,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00029509708737864075,
      "loss": 1.0283,
      "step": 223
    },
    {
      "epoch": 0.0905004797737488,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002950728155339805,
      "loss": 1.0652,
      "step": 224
    },
    {
      "epoch": 0.09090449977273875,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00029504854368932036,
      "loss": 1.0468,
      "step": 225
    },
    {
      "epoch": 0.0913085197717287,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002950242718446602,
      "loss": 1.0454,
      "step": 226
    },
    {
      "epoch": 0.09171253977071865,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00029499999999999996,
      "loss": 1.0093,
      "step": 227
    },
    {
      "epoch": 0.0921165597697086,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002949757281553398,
      "loss": 0.9629,
      "step": 228
    },
    {
      "epoch": 0.09252057976869855,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00029495145631067957,
      "loss": 0.9917,
      "step": 229
    },
    {
      "epoch": 0.0929245997676885,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002949271844660194,
      "loss": 1.1664,
      "step": 230
    },
    {
      "epoch": 0.09332861976667844,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0002949029126213592,
      "loss": 1.0614,
      "step": 231
    },
    {
      "epoch": 0.0937326397656684,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000294878640776699,
      "loss": 1.0827,
      "step": 232
    },
    {
      "epoch": 0.09413665976465835,
      "grad_norm": 0.8125,
      "learning_rate": 0.00029485436893203883,
      "loss": 0.8871,
      "step": 233
    },
    {
      "epoch": 0.0945406797636483,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0002948300970873786,
      "loss": 1.0594,
      "step": 234
    },
    {
      "epoch": 0.09494469976263825,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00029480582524271843,
      "loss": 1.0821,
      "step": 235
    },
    {
      "epoch": 0.0953487197616282,
      "grad_norm": 1.0,
      "learning_rate": 0.0002947815533980582,
      "loss": 0.9732,
      "step": 236
    },
    {
      "epoch": 0.09575273976061816,
      "grad_norm": 0.859375,
      "learning_rate": 0.00029475728155339804,
      "loss": 0.9982,
      "step": 237
    },
    {
      "epoch": 0.09615675975960811,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002947330097087378,
      "loss": 0.8946,
      "step": 238
    },
    {
      "epoch": 0.09656077975859804,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00029470873786407764,
      "loss": 1.0789,
      "step": 239
    },
    {
      "epoch": 0.096964799757588,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0002946844660194174,
      "loss": 1.0303,
      "step": 240
    },
    {
      "epoch": 0.09736881975657795,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00029466019417475725,
      "loss": 1.0468,
      "step": 241
    },
    {
      "epoch": 0.0977728397555679,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002946359223300971,
      "loss": 0.9786,
      "step": 242
    },
    {
      "epoch": 0.09817685975455785,
      "grad_norm": 0.77734375,
      "learning_rate": 0.00029461165048543685,
      "loss": 0.9,
      "step": 243
    },
    {
      "epoch": 0.0985808797535478,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0002945873786407767,
      "loss": 1.064,
      "step": 244
    },
    {
      "epoch": 0.09898489975253776,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00029456310679611646,
      "loss": 1.0181,
      "step": 245
    },
    {
      "epoch": 0.09938891975152771,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0002945388349514563,
      "loss": 1.0734,
      "step": 246
    },
    {
      "epoch": 0.09979293975051765,
      "grad_norm": 0.703125,
      "learning_rate": 0.0002945145631067961,
      "loss": 0.9849,
      "step": 247
    },
    {
      "epoch": 0.1001969597495076,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002944902912621359,
      "loss": 1.0185,
      "step": 248
    },
    {
      "epoch": 0.10060097974849755,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0002944660194174757,
      "loss": 0.9724,
      "step": 249
    },
    {
      "epoch": 0.1010049997474875,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002944417475728155,
      "loss": 0.9866,
      "step": 250
    },
    {
      "epoch": 0.10140901974647745,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0002944174757281553,
      "loss": 1.0369,
      "step": 251
    },
    {
      "epoch": 0.1018130397454674,
      "grad_norm": 0.6875,
      "learning_rate": 0.00029439320388349515,
      "loss": 0.936,
      "step": 252
    },
    {
      "epoch": 0.10221705974445736,
      "grad_norm": 0.84375,
      "learning_rate": 0.00029436893203883493,
      "loss": 1.1222,
      "step": 253
    },
    {
      "epoch": 0.1026210797434473,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0002943446601941747,
      "loss": 0.9793,
      "step": 254
    },
    {
      "epoch": 0.10302509974243725,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00029432038834951453,
      "loss": 0.8875,
      "step": 255
    },
    {
      "epoch": 0.1034291197414272,
      "grad_norm": 0.78125,
      "learning_rate": 0.00029429611650485436,
      "loss": 0.9942,
      "step": 256
    },
    {
      "epoch": 0.10383313974041715,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00029427184466019414,
      "loss": 0.9042,
      "step": 257
    },
    {
      "epoch": 0.1042371597394071,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00029424757281553397,
      "loss": 0.9496,
      "step": 258
    },
    {
      "epoch": 0.10464117973839705,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00029422330097087374,
      "loss": 0.9903,
      "step": 259
    },
    {
      "epoch": 0.105045199737387,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002941990291262136,
      "loss": 0.9087,
      "step": 260
    },
    {
      "epoch": 0.10544921973637696,
      "grad_norm": 0.703125,
      "learning_rate": 0.00029417475728155335,
      "loss": 0.9785,
      "step": 261
    },
    {
      "epoch": 0.1058532397353669,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0002941504854368932,
      "loss": 1.0158,
      "step": 262
    },
    {
      "epoch": 0.10625725973435685,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000294126213592233,
      "loss": 1.0027,
      "step": 263
    },
    {
      "epoch": 0.1066612797333468,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002941019417475728,
      "loss": 1.1421,
      "step": 264
    },
    {
      "epoch": 0.10706529973233675,
      "grad_norm": 0.75390625,
      "learning_rate": 0.00029407766990291256,
      "loss": 0.8618,
      "step": 265
    },
    {
      "epoch": 0.1074693197313267,
      "grad_norm": 0.75,
      "learning_rate": 0.0002940533980582524,
      "loss": 0.9511,
      "step": 266
    },
    {
      "epoch": 0.10787333973031665,
      "grad_norm": 2.96875,
      "learning_rate": 0.0002940291262135922,
      "loss": 0.8486,
      "step": 267
    },
    {
      "epoch": 0.1082773597293066,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000294004854368932,
      "loss": 0.9963,
      "step": 268
    },
    {
      "epoch": 0.10868137972829656,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0002939805825242718,
      "loss": 0.9622,
      "step": 269
    },
    {
      "epoch": 0.1090853997272865,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002939563106796116,
      "loss": 0.9933,
      "step": 270
    },
    {
      "epoch": 0.10948941972627645,
      "grad_norm": 0.70703125,
      "learning_rate": 0.0002939320388349514,
      "loss": 0.8924,
      "step": 271
    },
    {
      "epoch": 0.1098934397252664,
      "grad_norm": 0.7421875,
      "learning_rate": 0.00029390776699029126,
      "loss": 0.8703,
      "step": 272
    },
    {
      "epoch": 0.11029745972425635,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00029388349514563103,
      "loss": 0.8987,
      "step": 273
    },
    {
      "epoch": 0.1107014797232463,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00029385922330097086,
      "loss": 1.0079,
      "step": 274
    },
    {
      "epoch": 0.11110549972223625,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00029383495145631064,
      "loss": 1.112,
      "step": 275
    },
    {
      "epoch": 0.1115095197212262,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00029381067961165047,
      "loss": 0.9664,
      "step": 276
    },
    {
      "epoch": 0.11191353972021616,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0002937864077669903,
      "loss": 0.8644,
      "step": 277
    },
    {
      "epoch": 0.1123175597192061,
      "grad_norm": 0.76953125,
      "learning_rate": 0.00029376213592233007,
      "loss": 0.9472,
      "step": 278
    },
    {
      "epoch": 0.11272157971819605,
      "grad_norm": 0.8125,
      "learning_rate": 0.0002937378640776699,
      "loss": 0.9776,
      "step": 279
    },
    {
      "epoch": 0.113125599717186,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0002937135922330097,
      "loss": 1.0549,
      "step": 280
    },
    {
      "epoch": 0.11352961971617595,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0002936893203883495,
      "loss": 0.9088,
      "step": 281
    },
    {
      "epoch": 0.1139336397151659,
      "grad_norm": 0.78515625,
      "learning_rate": 0.00029366504854368933,
      "loss": 1.0217,
      "step": 282
    },
    {
      "epoch": 0.11433765971415585,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002936407766990291,
      "loss": 0.9649,
      "step": 283
    },
    {
      "epoch": 0.1147416797131458,
      "grad_norm": 3.28125,
      "learning_rate": 0.0002936165048543689,
      "loss": 1.1714,
      "step": 284
    },
    {
      "epoch": 0.11514569971213576,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0002935922330097087,
      "loss": 1.1016,
      "step": 285
    },
    {
      "epoch": 0.1155497197111257,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002935679611650485,
      "loss": 0.9771,
      "step": 286
    },
    {
      "epoch": 0.11595373971011565,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002935436893203883,
      "loss": 0.9563,
      "step": 287
    },
    {
      "epoch": 0.1163577597091056,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00029351941747572815,
      "loss": 0.9222,
      "step": 288
    },
    {
      "epoch": 0.11676177970809555,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002934951456310679,
      "loss": 0.9364,
      "step": 289
    },
    {
      "epoch": 0.1171657997070855,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00029347087378640775,
      "loss": 1.0015,
      "step": 290
    },
    {
      "epoch": 0.11756981970607545,
      "grad_norm": 0.703125,
      "learning_rate": 0.00029344660194174753,
      "loss": 0.9835,
      "step": 291
    },
    {
      "epoch": 0.1179738397050654,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00029342233009708736,
      "loss": 1.0738,
      "step": 292
    },
    {
      "epoch": 0.11837785970405536,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002933980582524272,
      "loss": 1.006,
      "step": 293
    },
    {
      "epoch": 0.1187818797030453,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029337378640776696,
      "loss": 0.9796,
      "step": 294
    },
    {
      "epoch": 0.11918589970203525,
      "grad_norm": 3.296875,
      "learning_rate": 0.00029334951456310674,
      "loss": 1.0246,
      "step": 295
    },
    {
      "epoch": 0.1195899197010252,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00029332524271844657,
      "loss": 0.9388,
      "step": 296
    },
    {
      "epoch": 0.11999393970001515,
      "grad_norm": 0.765625,
      "learning_rate": 0.0002933009708737864,
      "loss": 1.0142,
      "step": 297
    },
    {
      "epoch": 0.1203979596990051,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00029327669902912617,
      "loss": 1.0849,
      "step": 298
    },
    {
      "epoch": 0.12080197969799505,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000293252427184466,
      "loss": 0.9702,
      "step": 299
    },
    {
      "epoch": 0.121205999696985,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0002932281553398058,
      "loss": 1.1123,
      "step": 300
    },
    {
      "epoch": 0.12161001969597494,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0002932038834951456,
      "loss": 0.9149,
      "step": 301
    },
    {
      "epoch": 0.1220140396949649,
      "grad_norm": 0.8125,
      "learning_rate": 0.00029317961165048543,
      "loss": 0.9886,
      "step": 302
    },
    {
      "epoch": 0.12241805969395485,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002931553398058252,
      "loss": 0.9678,
      "step": 303
    },
    {
      "epoch": 0.1228220796929448,
      "grad_norm": 0.75,
      "learning_rate": 0.00029313106796116504,
      "loss": 0.8933,
      "step": 304
    },
    {
      "epoch": 0.12322609969193475,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0002931067961165048,
      "loss": 1.0086,
      "step": 305
    },
    {
      "epoch": 0.1236301196909247,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00029308252427184464,
      "loss": 0.9656,
      "step": 306
    },
    {
      "epoch": 0.12403413968991465,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0002930582524271845,
      "loss": 1.0638,
      "step": 307
    },
    {
      "epoch": 0.1244381596889046,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00029303398058252425,
      "loss": 0.9599,
      "step": 308
    },
    {
      "epoch": 0.12484217968789454,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0002930097087378641,
      "loss": 1.0094,
      "step": 309
    },
    {
      "epoch": 0.1252461996868845,
      "grad_norm": 0.76953125,
      "learning_rate": 0.00029298543689320385,
      "loss": 0.9661,
      "step": 310
    },
    {
      "epoch": 0.12565021968587445,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002929611650485437,
      "loss": 0.865,
      "step": 311
    },
    {
      "epoch": 0.1260542396848644,
      "grad_norm": 0.828125,
      "learning_rate": 0.00029293689320388346,
      "loss": 1.0541,
      "step": 312
    },
    {
      "epoch": 0.12645825968385435,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002929126213592233,
      "loss": 0.8174,
      "step": 313
    },
    {
      "epoch": 0.1268622796828443,
      "grad_norm": 0.734375,
      "learning_rate": 0.00029288834951456306,
      "loss": 1.0021,
      "step": 314
    },
    {
      "epoch": 0.12726629968183426,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002928640776699029,
      "loss": 0.8684,
      "step": 315
    },
    {
      "epoch": 0.1276703196808242,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00029283980582524267,
      "loss": 0.9502,
      "step": 316
    },
    {
      "epoch": 0.12807433967981416,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002928155339805825,
      "loss": 1.0258,
      "step": 317
    },
    {
      "epoch": 0.1284783596788041,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0002927912621359223,
      "loss": 0.998,
      "step": 318
    },
    {
      "epoch": 0.12888237967779406,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0002927669902912621,
      "loss": 1.1369,
      "step": 319
    },
    {
      "epoch": 0.12928639967678401,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00029274271844660193,
      "loss": 1.0052,
      "step": 320
    },
    {
      "epoch": 0.12969041967577394,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002927184466019417,
      "loss": 1.0297,
      "step": 321
    },
    {
      "epoch": 0.1300944396747639,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00029269417475728154,
      "loss": 0.8447,
      "step": 322
    },
    {
      "epoch": 0.13049845967375384,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00029266990291262136,
      "loss": 0.8856,
      "step": 323
    },
    {
      "epoch": 0.1309024796727438,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00029264563106796114,
      "loss": 1.0237,
      "step": 324
    },
    {
      "epoch": 0.13130649967173375,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0002926213592233009,
      "loss": 0.9483,
      "step": 325
    },
    {
      "epoch": 0.1317105196707237,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00029259708737864074,
      "loss": 0.9272,
      "step": 326
    },
    {
      "epoch": 0.13211453966971365,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002925728155339806,
      "loss": 0.9646,
      "step": 327
    },
    {
      "epoch": 0.1325185596687036,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00029254854368932035,
      "loss": 0.9909,
      "step": 328
    },
    {
      "epoch": 0.13292257966769355,
      "grad_norm": 0.765625,
      "learning_rate": 0.0002925242718446602,
      "loss": 1.0682,
      "step": 329
    },
    {
      "epoch": 0.1333265996666835,
      "grad_norm": 0.71875,
      "learning_rate": 0.00029249999999999995,
      "loss": 1.0289,
      "step": 330
    },
    {
      "epoch": 0.13373061966567346,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0002924757281553398,
      "loss": 1.1211,
      "step": 331
    },
    {
      "epoch": 0.1341346396646634,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002924514563106796,
      "loss": 0.9447,
      "step": 332
    },
    {
      "epoch": 0.13453865966365336,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002924271844660194,
      "loss": 0.9128,
      "step": 333
    },
    {
      "epoch": 0.1349426796626433,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0002924029126213592,
      "loss": 0.9529,
      "step": 334
    },
    {
      "epoch": 0.13534669966163326,
      "grad_norm": 0.703125,
      "learning_rate": 0.000292378640776699,
      "loss": 0.92,
      "step": 335
    },
    {
      "epoch": 0.13575071966062321,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0002923543689320388,
      "loss": 0.9382,
      "step": 336
    },
    {
      "epoch": 0.13615473965961314,
      "grad_norm": 0.78125,
      "learning_rate": 0.0002923300970873786,
      "loss": 0.9448,
      "step": 337
    },
    {
      "epoch": 0.1365587596586031,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0002923058252427184,
      "loss": 1.0835,
      "step": 338
    },
    {
      "epoch": 0.13696277965759304,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002922815533980582,
      "loss": 0.9923,
      "step": 339
    },
    {
      "epoch": 0.137366799656583,
      "grad_norm": 0.78125,
      "learning_rate": 0.00029225728155339803,
      "loss": 1.0372,
      "step": 340
    },
    {
      "epoch": 0.13777081965557295,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002922330097087378,
      "loss": 0.8467,
      "step": 341
    },
    {
      "epoch": 0.1381748396545629,
      "grad_norm": 0.71875,
      "learning_rate": 0.00029220873786407764,
      "loss": 1.0142,
      "step": 342
    },
    {
      "epoch": 0.13857885965355285,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00029218446601941747,
      "loss": 0.9025,
      "step": 343
    },
    {
      "epoch": 0.1389828796525428,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00029216019417475724,
      "loss": 0.921,
      "step": 344
    },
    {
      "epoch": 0.13938689965153275,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00029213592233009707,
      "loss": 1.0033,
      "step": 345
    },
    {
      "epoch": 0.1397909196505227,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00029211165048543685,
      "loss": 0.8607,
      "step": 346
    },
    {
      "epoch": 0.14019493964951266,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002920873786407767,
      "loss": 0.9075,
      "step": 347
    },
    {
      "epoch": 0.1405989596485026,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0002920631067961165,
      "loss": 1.0099,
      "step": 348
    },
    {
      "epoch": 0.14100297964749256,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002920388349514563,
      "loss": 0.9787,
      "step": 349
    },
    {
      "epoch": 0.1414069996464825,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002920145631067961,
      "loss": 1.046,
      "step": 350
    },
    {
      "epoch": 0.14181101964547246,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0002919902912621359,
      "loss": 0.95,
      "step": 351
    },
    {
      "epoch": 0.1422150396444624,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002919660194174757,
      "loss": 0.9394,
      "step": 352
    },
    {
      "epoch": 0.14261905964345234,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00029194174757281554,
      "loss": 1.0869,
      "step": 353
    },
    {
      "epoch": 0.1430230796424423,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002919174757281553,
      "loss": 0.8593,
      "step": 354
    },
    {
      "epoch": 0.14342709964143224,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0002918932038834951,
      "loss": 0.9251,
      "step": 355
    },
    {
      "epoch": 0.1438311196404222,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002918689320388349,
      "loss": 1.0662,
      "step": 356
    },
    {
      "epoch": 0.14423513963941215,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00029184466019417475,
      "loss": 0.9809,
      "step": 357
    },
    {
      "epoch": 0.1446391596384021,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00029182038834951453,
      "loss": 0.9219,
      "step": 358
    },
    {
      "epoch": 0.14504317963739205,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00029179611650485436,
      "loss": 0.9277,
      "step": 359
    },
    {
      "epoch": 0.145447199636382,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00029177184466019413,
      "loss": 0.9504,
      "step": 360
    },
    {
      "epoch": 0.14585121963537195,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00029174757281553396,
      "loss": 0.947,
      "step": 361
    },
    {
      "epoch": 0.1462552396343619,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00029172330097087374,
      "loss": 0.978,
      "step": 362
    },
    {
      "epoch": 0.14665925963335186,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00029169902912621357,
      "loss": 0.8699,
      "step": 363
    },
    {
      "epoch": 0.1470632796323418,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002916747572815534,
      "loss": 1.1887,
      "step": 364
    },
    {
      "epoch": 0.14746729963133176,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029165048543689317,
      "loss": 1.0645,
      "step": 365
    },
    {
      "epoch": 0.1478713196303217,
      "grad_norm": 0.76953125,
      "learning_rate": 0.00029162621359223295,
      "loss": 0.9479,
      "step": 366
    },
    {
      "epoch": 0.14827533962931166,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0002916019417475728,
      "loss": 1.1128,
      "step": 367
    },
    {
      "epoch": 0.1486793596283016,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002915776699029126,
      "loss": 0.8824,
      "step": 368
    },
    {
      "epoch": 0.14908337962729154,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0002915533980582524,
      "loss": 1.0144,
      "step": 369
    },
    {
      "epoch": 0.1494873996262815,
      "grad_norm": 0.78125,
      "learning_rate": 0.0002915291262135922,
      "loss": 0.9871,
      "step": 370
    },
    {
      "epoch": 0.14989141962527144,
      "grad_norm": 0.640625,
      "learning_rate": 0.000291504854368932,
      "loss": 0.9508,
      "step": 371
    },
    {
      "epoch": 0.1502954396242614,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0002914805825242718,
      "loss": 0.9727,
      "step": 372
    },
    {
      "epoch": 0.15069945962325135,
      "grad_norm": 0.796875,
      "learning_rate": 0.00029145631067961164,
      "loss": 0.9153,
      "step": 373
    },
    {
      "epoch": 0.1511034796222413,
      "grad_norm": 0.671875,
      "learning_rate": 0.0002914320388349514,
      "loss": 0.9147,
      "step": 374
    },
    {
      "epoch": 0.15150749962123125,
      "grad_norm": 0.6875,
      "learning_rate": 0.00029140776699029125,
      "loss": 0.9147,
      "step": 375
    },
    {
      "epoch": 0.1519115196202212,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000291383495145631,
      "loss": 1.018,
      "step": 376
    },
    {
      "epoch": 0.15231553961921115,
      "grad_norm": 0.71875,
      "learning_rate": 0.00029135922330097085,
      "loss": 0.8901,
      "step": 377
    },
    {
      "epoch": 0.1527195596182011,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002913349514563107,
      "loss": 0.9848,
      "step": 378
    },
    {
      "epoch": 0.15312357961719106,
      "grad_norm": 0.75,
      "learning_rate": 0.00029131067961165046,
      "loss": 1.0765,
      "step": 379
    },
    {
      "epoch": 0.153527599616181,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002912864077669903,
      "loss": 0.9138,
      "step": 380
    },
    {
      "epoch": 0.15393161961517096,
      "grad_norm": 0.78125,
      "learning_rate": 0.00029126213592233006,
      "loss": 0.8982,
      "step": 381
    },
    {
      "epoch": 0.1543356396141609,
      "grad_norm": 0.671875,
      "learning_rate": 0.0002912378640776699,
      "loss": 0.9418,
      "step": 382
    },
    {
      "epoch": 0.15473965961315084,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0002912135922330097,
      "loss": 0.9668,
      "step": 383
    },
    {
      "epoch": 0.1551436796121408,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002911893203883495,
      "loss": 0.9175,
      "step": 384
    },
    {
      "epoch": 0.15554769961113074,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00029116504854368927,
      "loss": 1.0371,
      "step": 385
    },
    {
      "epoch": 0.1559517196101207,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0002911407766990291,
      "loss": 1.0674,
      "step": 386
    },
    {
      "epoch": 0.15635573960911064,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0002911165048543689,
      "loss": 1.0343,
      "step": 387
    },
    {
      "epoch": 0.1567597596081006,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002910922330097087,
      "loss": 0.9558,
      "step": 388
    },
    {
      "epoch": 0.15716377960709055,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00029106796116504854,
      "loss": 1.0743,
      "step": 389
    },
    {
      "epoch": 0.1575677996060805,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002910436893203883,
      "loss": 1.0006,
      "step": 390
    },
    {
      "epoch": 0.15797181960507045,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00029101941747572814,
      "loss": 1.0064,
      "step": 391
    },
    {
      "epoch": 0.1583758396040604,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002909951456310679,
      "loss": 1.0054,
      "step": 392
    },
    {
      "epoch": 0.15877985960305036,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00029097087378640775,
      "loss": 0.821,
      "step": 393
    },
    {
      "epoch": 0.1591838796020403,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002909466019417476,
      "loss": 0.9554,
      "step": 394
    },
    {
      "epoch": 0.15958789960103026,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00029092233009708735,
      "loss": 0.8887,
      "step": 395
    },
    {
      "epoch": 0.1599919196000202,
      "grad_norm": 0.75,
      "learning_rate": 0.0002908980582524271,
      "loss": 1.0168,
      "step": 396
    },
    {
      "epoch": 0.16039593959901016,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00029087378640776695,
      "loss": 0.9522,
      "step": 397
    },
    {
      "epoch": 0.16079995959800011,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0002908495145631068,
      "loss": 0.96,
      "step": 398
    },
    {
      "epoch": 0.16120397959699004,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00029082524271844656,
      "loss": 0.9688,
      "step": 399
    },
    {
      "epoch": 0.16160799959598,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002908009708737864,
      "loss": 0.9729,
      "step": 400
    },
    {
      "epoch": 0.16201201959496994,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00029077669902912616,
      "loss": 0.9569,
      "step": 401
    },
    {
      "epoch": 0.1624160395939599,
      "grad_norm": 0.69921875,
      "learning_rate": 0.000290752427184466,
      "loss": 0.9364,
      "step": 402
    },
    {
      "epoch": 0.16282005959294985,
      "grad_norm": 0.625,
      "learning_rate": 0.0002907281553398058,
      "loss": 0.8842,
      "step": 403
    },
    {
      "epoch": 0.1632240795919398,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002907038834951456,
      "loss": 0.8996,
      "step": 404
    },
    {
      "epoch": 0.16362809959092975,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00029067961165048543,
      "loss": 0.8592,
      "step": 405
    },
    {
      "epoch": 0.1640321195899197,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002906553398058252,
      "loss": 0.946,
      "step": 406
    },
    {
      "epoch": 0.16443613958890965,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00029063106796116503,
      "loss": 0.9374,
      "step": 407
    },
    {
      "epoch": 0.1648401595878996,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00029060679611650486,
      "loss": 0.9407,
      "step": 408
    },
    {
      "epoch": 0.16524417958688956,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00029058252427184464,
      "loss": 0.9434,
      "step": 409
    },
    {
      "epoch": 0.1656481995858795,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00029055825242718447,
      "loss": 0.8999,
      "step": 410
    },
    {
      "epoch": 0.16605221958486946,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029053398058252424,
      "loss": 0.9597,
      "step": 411
    },
    {
      "epoch": 0.1664562395838594,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00029050970873786407,
      "loss": 0.9277,
      "step": 412
    },
    {
      "epoch": 0.16686025958284936,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00029048543689320385,
      "loss": 0.9732,
      "step": 413
    },
    {
      "epoch": 0.16726427958183931,
      "grad_norm": 0.625,
      "learning_rate": 0.0002904611650485437,
      "loss": 0.9582,
      "step": 414
    },
    {
      "epoch": 0.16766829958082924,
      "grad_norm": 0.609375,
      "learning_rate": 0.00029043689320388345,
      "loss": 0.8973,
      "step": 415
    },
    {
      "epoch": 0.1680723195798192,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002904126213592233,
      "loss": 0.9233,
      "step": 416
    },
    {
      "epoch": 0.16847633957880914,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00029038834951456306,
      "loss": 1.0908,
      "step": 417
    },
    {
      "epoch": 0.1688803595777991,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0002903640776699029,
      "loss": 0.9227,
      "step": 418
    },
    {
      "epoch": 0.16928437957678905,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002903398058252427,
      "loss": 0.8926,
      "step": 419
    },
    {
      "epoch": 0.169688399575779,
      "grad_norm": 0.859375,
      "learning_rate": 0.0002903155339805825,
      "loss": 1.0873,
      "step": 420
    },
    {
      "epoch": 0.17009241957476895,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002902912621359223,
      "loss": 0.8927,
      "step": 421
    },
    {
      "epoch": 0.1704964395737589,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002902669902912621,
      "loss": 0.9399,
      "step": 422
    },
    {
      "epoch": 0.17090045957274885,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002902427184466019,
      "loss": 0.9971,
      "step": 423
    },
    {
      "epoch": 0.1713044795717388,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00029021844660194175,
      "loss": 0.9264,
      "step": 424
    },
    {
      "epoch": 0.17170849957072876,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00029019417475728153,
      "loss": 0.8329,
      "step": 425
    },
    {
      "epoch": 0.1721125195697187,
      "grad_norm": 0.70703125,
      "learning_rate": 0.0002901699029126213,
      "loss": 0.9387,
      "step": 426
    },
    {
      "epoch": 0.17251653956870866,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00029014563106796113,
      "loss": 0.9553,
      "step": 427
    },
    {
      "epoch": 0.1729205595676986,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00029012135922330096,
      "loss": 1.0072,
      "step": 428
    },
    {
      "epoch": 0.17332457956668856,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00029009708737864074,
      "loss": 1.0054,
      "step": 429
    },
    {
      "epoch": 0.1737285995656785,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00029007281553398057,
      "loss": 0.7583,
      "step": 430
    },
    {
      "epoch": 0.17413261956466844,
      "grad_norm": 0.640625,
      "learning_rate": 0.00029004854368932034,
      "loss": 0.8443,
      "step": 431
    },
    {
      "epoch": 0.1745366395636584,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00029002427184466017,
      "loss": 0.9096,
      "step": 432
    },
    {
      "epoch": 0.17494065956264834,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00029,
      "loss": 0.8336,
      "step": 433
    },
    {
      "epoch": 0.1753446795616383,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002899757281553398,
      "loss": 0.8841,
      "step": 434
    },
    {
      "epoch": 0.17574869956062825,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002899514563106796,
      "loss": 1.0032,
      "step": 435
    },
    {
      "epoch": 0.1761527195596182,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0002899271844660194,
      "loss": 0.9524,
      "step": 436
    },
    {
      "epoch": 0.17655673955860815,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002899029126213592,
      "loss": 0.8574,
      "step": 437
    },
    {
      "epoch": 0.1769607595575981,
      "grad_norm": 1.015625,
      "learning_rate": 0.000289878640776699,
      "loss": 1.0049,
      "step": 438
    },
    {
      "epoch": 0.17736477955658805,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0002898543689320388,
      "loss": 1.0577,
      "step": 439
    },
    {
      "epoch": 0.177768799555578,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0002898300970873786,
      "loss": 0.9255,
      "step": 440
    },
    {
      "epoch": 0.17817281955456796,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002898058252427184,
      "loss": 0.8805,
      "step": 441
    },
    {
      "epoch": 0.1785768395535579,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002897815533980582,
      "loss": 1.116,
      "step": 442
    },
    {
      "epoch": 0.17898085955254786,
      "grad_norm": 0.66796875,
      "learning_rate": 0.000289757281553398,
      "loss": 0.8507,
      "step": 443
    },
    {
      "epoch": 0.1793848795515378,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00028973300970873785,
      "loss": 0.9159,
      "step": 444
    },
    {
      "epoch": 0.17978889955052776,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00028970873786407763,
      "loss": 0.9899,
      "step": 445
    },
    {
      "epoch": 0.1801929195495177,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00028968446601941746,
      "loss": 1.0286,
      "step": 446
    },
    {
      "epoch": 0.18059693954850764,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00028966019417475723,
      "loss": 1.0754,
      "step": 447
    },
    {
      "epoch": 0.1810009595474976,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00028963592233009706,
      "loss": 0.9649,
      "step": 448
    },
    {
      "epoch": 0.18140497954648754,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002896116504854369,
      "loss": 0.9307,
      "step": 449
    },
    {
      "epoch": 0.1818089995454775,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00028958737864077667,
      "loss": 1.0451,
      "step": 450
    },
    {
      "epoch": 0.18221301954446745,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002895631067961165,
      "loss": 1.0017,
      "step": 451
    },
    {
      "epoch": 0.1826170395434574,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002895388349514563,
      "loss": 0.8848,
      "step": 452
    },
    {
      "epoch": 0.18302105954244735,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0002895145631067961,
      "loss": 0.9599,
      "step": 453
    },
    {
      "epoch": 0.1834250795414373,
      "grad_norm": 0.78125,
      "learning_rate": 0.00028949029126213593,
      "loss": 1.0509,
      "step": 454
    },
    {
      "epoch": 0.18382909954042725,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002894660194174757,
      "loss": 0.9038,
      "step": 455
    },
    {
      "epoch": 0.1842331195394172,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002894417475728155,
      "loss": 0.9076,
      "step": 456
    },
    {
      "epoch": 0.18463713953840716,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002894174757281553,
      "loss": 0.8815,
      "step": 457
    },
    {
      "epoch": 0.1850411595373971,
      "grad_norm": 0.76171875,
      "learning_rate": 0.00028939320388349514,
      "loss": 1.1018,
      "step": 458
    },
    {
      "epoch": 0.18544517953638706,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002893689320388349,
      "loss": 0.8771,
      "step": 459
    },
    {
      "epoch": 0.185849199535377,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00028934466019417475,
      "loss": 0.9859,
      "step": 460
    },
    {
      "epoch": 0.18625321953436696,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0002893203883495145,
      "loss": 1.0455,
      "step": 461
    },
    {
      "epoch": 0.1866572395333569,
      "grad_norm": 0.546875,
      "learning_rate": 0.00028929611650485435,
      "loss": 0.8593,
      "step": 462
    },
    {
      "epoch": 0.18706125953234684,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002892718446601941,
      "loss": 0.9824,
      "step": 463
    },
    {
      "epoch": 0.1874652795313368,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00028924757281553396,
      "loss": 0.9396,
      "step": 464
    },
    {
      "epoch": 0.18786929953032674,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002892233009708738,
      "loss": 0.9899,
      "step": 465
    },
    {
      "epoch": 0.1882733195293167,
      "grad_norm": 0.6875,
      "learning_rate": 0.00028919902912621356,
      "loss": 1.0053,
      "step": 466
    },
    {
      "epoch": 0.18867733952830665,
      "grad_norm": 0.6875,
      "learning_rate": 0.00028917475728155334,
      "loss": 0.9602,
      "step": 467
    },
    {
      "epoch": 0.1890813595272966,
      "grad_norm": 0.71875,
      "learning_rate": 0.00028915048543689316,
      "loss": 0.9674,
      "step": 468
    },
    {
      "epoch": 0.18948537952628655,
      "grad_norm": 0.6484375,
      "learning_rate": 0.000289126213592233,
      "loss": 0.9548,
      "step": 469
    },
    {
      "epoch": 0.1898893995252765,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00028910194174757277,
      "loss": 0.9667,
      "step": 470
    },
    {
      "epoch": 0.19029341952426646,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002890776699029126,
      "loss": 0.9601,
      "step": 471
    },
    {
      "epoch": 0.1906974395232564,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002890533980582524,
      "loss": 1.0437,
      "step": 472
    },
    {
      "epoch": 0.19110145952224636,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0002890291262135922,
      "loss": 1.013,
      "step": 473
    },
    {
      "epoch": 0.1915054795212363,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00028900485436893203,
      "loss": 0.9955,
      "step": 474
    },
    {
      "epoch": 0.19190949952022626,
      "grad_norm": 6.5,
      "learning_rate": 0.0002889805825242718,
      "loss": 1.217,
      "step": 475
    },
    {
      "epoch": 0.19231351951921621,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00028895631067961164,
      "loss": 0.9331,
      "step": 476
    },
    {
      "epoch": 0.19271753951820614,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0002889320388349514,
      "loss": 0.9918,
      "step": 477
    },
    {
      "epoch": 0.1931215595171961,
      "grad_norm": 0.703125,
      "learning_rate": 0.00028890776699029124,
      "loss": 0.9231,
      "step": 478
    },
    {
      "epoch": 0.19352557951618604,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00028888349514563107,
      "loss": 0.8452,
      "step": 479
    },
    {
      "epoch": 0.193929599515176,
      "grad_norm": 0.625,
      "learning_rate": 0.00028885922330097085,
      "loss": 0.9044,
      "step": 480
    },
    {
      "epoch": 0.19433361951416595,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002888349514563107,
      "loss": 0.9561,
      "step": 481
    },
    {
      "epoch": 0.1947376395131559,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00028881067961165045,
      "loss": 0.9036,
      "step": 482
    },
    {
      "epoch": 0.19514165951214585,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002887864077669903,
      "loss": 0.8151,
      "step": 483
    },
    {
      "epoch": 0.1955456795111358,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002887621359223301,
      "loss": 0.8803,
      "step": 484
    },
    {
      "epoch": 0.19594969951012575,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0002887378640776699,
      "loss": 0.9758,
      "step": 485
    },
    {
      "epoch": 0.1963537195091157,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00028871359223300966,
      "loss": 0.8342,
      "step": 486
    },
    {
      "epoch": 0.19675773950810566,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0002886893203883495,
      "loss": 0.9107,
      "step": 487
    },
    {
      "epoch": 0.1971617595070956,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00028866504854368927,
      "loss": 0.937,
      "step": 488
    },
    {
      "epoch": 0.19756577950608556,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0002886407766990291,
      "loss": 0.9575,
      "step": 489
    },
    {
      "epoch": 0.1979697995050755,
      "grad_norm": 0.796875,
      "learning_rate": 0.0002886165048543689,
      "loss": 0.9534,
      "step": 490
    },
    {
      "epoch": 0.19837381950406546,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0002885922330097087,
      "loss": 0.9845,
      "step": 491
    },
    {
      "epoch": 0.19877783950305541,
      "grad_norm": 0.78125,
      "learning_rate": 0.00028856796116504853,
      "loss": 1.0751,
      "step": 492
    },
    {
      "epoch": 0.19918185950204534,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0002885436893203883,
      "loss": 0.9038,
      "step": 493
    },
    {
      "epoch": 0.1995858795010353,
      "grad_norm": 0.6875,
      "learning_rate": 0.00028851941747572813,
      "loss": 0.9239,
      "step": 494
    },
    {
      "epoch": 0.19998989950002524,
      "grad_norm": 0.671875,
      "learning_rate": 0.00028849514563106796,
      "loss": 0.9502,
      "step": 495
    },
    {
      "epoch": 0.2003939194990152,
      "grad_norm": 0.609375,
      "learning_rate": 0.00028847087378640774,
      "loss": 0.8342,
      "step": 496
    },
    {
      "epoch": 0.20079793949800515,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002884466019417475,
      "loss": 0.8999,
      "step": 497
    },
    {
      "epoch": 0.2012019594969951,
      "grad_norm": 0.71875,
      "learning_rate": 0.00028842233009708734,
      "loss": 1.0026,
      "step": 498
    },
    {
      "epoch": 0.20160597949598505,
      "grad_norm": 0.625,
      "learning_rate": 0.0002883980582524272,
      "loss": 0.957,
      "step": 499
    },
    {
      "epoch": 0.202009999494975,
      "grad_norm": 0.609375,
      "learning_rate": 0.00028837378640776695,
      "loss": 0.8801,
      "step": 500
    },
    {
      "epoch": 0.20241401949396495,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0002883495145631068,
      "loss": 0.9967,
      "step": 501
    },
    {
      "epoch": 0.2028180394929549,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00028832524271844655,
      "loss": 0.9743,
      "step": 502
    },
    {
      "epoch": 0.20322205949194486,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002883009708737864,
      "loss": 1.0198,
      "step": 503
    },
    {
      "epoch": 0.2036260794909348,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0002882766990291262,
      "loss": 0.9923,
      "step": 504
    },
    {
      "epoch": 0.20403009948992476,
      "grad_norm": 0.75390625,
      "learning_rate": 0.000288252427184466,
      "loss": 1.0495,
      "step": 505
    },
    {
      "epoch": 0.2044341194889147,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0002882281553398058,
      "loss": 1.0756,
      "step": 506
    },
    {
      "epoch": 0.20483813948790466,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002882038834951456,
      "loss": 0.8983,
      "step": 507
    },
    {
      "epoch": 0.2052421594868946,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0002881796116504854,
      "loss": 0.9395,
      "step": 508
    },
    {
      "epoch": 0.20564617948588454,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00028815533980582525,
      "loss": 0.8725,
      "step": 509
    },
    {
      "epoch": 0.2060501994848745,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000288131067961165,
      "loss": 1.0873,
      "step": 510
    },
    {
      "epoch": 0.20645421948386444,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00028810679611650486,
      "loss": 0.9895,
      "step": 511
    },
    {
      "epoch": 0.2068582394828544,
      "grad_norm": 0.734375,
      "learning_rate": 0.00028808252427184463,
      "loss": 0.884,
      "step": 512
    },
    {
      "epoch": 0.20726225948184435,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002880582524271844,
      "loss": 0.869,
      "step": 513
    },
    {
      "epoch": 0.2076662794808343,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00028803398058252424,
      "loss": 0.9871,
      "step": 514
    },
    {
      "epoch": 0.20807029947982425,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00028800970873786406,
      "loss": 0.8309,
      "step": 515
    },
    {
      "epoch": 0.2084743194788142,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00028798543689320384,
      "loss": 1.0642,
      "step": 516
    },
    {
      "epoch": 0.20887833947780415,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00028796116504854367,
      "loss": 0.9013,
      "step": 517
    },
    {
      "epoch": 0.2092823594767941,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00028793689320388344,
      "loss": 1.138,
      "step": 518
    },
    {
      "epoch": 0.20968637947578406,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0002879126213592233,
      "loss": 0.8747,
      "step": 519
    },
    {
      "epoch": 0.210090399474774,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002878883495145631,
      "loss": 1.0327,
      "step": 520
    },
    {
      "epoch": 0.21049441947376396,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0002878640776699029,
      "loss": 0.9591,
      "step": 521
    },
    {
      "epoch": 0.2108984394727539,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002878398058252427,
      "loss": 0.9316,
      "step": 522
    },
    {
      "epoch": 0.21130245947174386,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0002878155339805825,
      "loss": 1.0135,
      "step": 523
    },
    {
      "epoch": 0.2117064794707338,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0002877912621359223,
      "loss": 0.884,
      "step": 524
    },
    {
      "epoch": 0.21211049946972374,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00028776699029126214,
      "loss": 1.0663,
      "step": 525
    },
    {
      "epoch": 0.2125145194687137,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002877427184466019,
      "loss": 0.892,
      "step": 526
    },
    {
      "epoch": 0.21291853946770364,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002877184466019417,
      "loss": 1.0309,
      "step": 527
    },
    {
      "epoch": 0.2133225594666936,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002876941747572815,
      "loss": 0.8718,
      "step": 528
    },
    {
      "epoch": 0.21372657946568355,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00028766990291262135,
      "loss": 0.983,
      "step": 529
    },
    {
      "epoch": 0.2141305994646735,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002876456310679611,
      "loss": 0.9089,
      "step": 530
    },
    {
      "epoch": 0.21453461946366345,
      "grad_norm": 0.703125,
      "learning_rate": 0.00028762135922330096,
      "loss": 1.0427,
      "step": 531
    },
    {
      "epoch": 0.2149386394626534,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00028759708737864073,
      "loss": 0.9013,
      "step": 532
    },
    {
      "epoch": 0.21534265946164335,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00028757281553398056,
      "loss": 0.8577,
      "step": 533
    },
    {
      "epoch": 0.2157466794606333,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002875485436893204,
      "loss": 0.896,
      "step": 534
    },
    {
      "epoch": 0.21615069945962326,
      "grad_norm": 0.640625,
      "learning_rate": 0.00028752427184466017,
      "loss": 0.8859,
      "step": 535
    },
    {
      "epoch": 0.2165547194586132,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0002875,
      "loss": 0.8967,
      "step": 536
    },
    {
      "epoch": 0.21695873945760316,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00028747572815533977,
      "loss": 0.9091,
      "step": 537
    },
    {
      "epoch": 0.2173627594565931,
      "grad_norm": 0.625,
      "learning_rate": 0.0002874514563106796,
      "loss": 0.9385,
      "step": 538
    },
    {
      "epoch": 0.21776677945558306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002874271844660194,
      "loss": 1.0022,
      "step": 539
    },
    {
      "epoch": 0.218170799454573,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002874029126213592,
      "loss": 0.8623,
      "step": 540
    },
    {
      "epoch": 0.21857481945356294,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000287378640776699,
      "loss": 1.0123,
      "step": 541
    },
    {
      "epoch": 0.2189788394525529,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002873543689320388,
      "loss": 0.9064,
      "step": 542
    },
    {
      "epoch": 0.21938285945154284,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002873300970873786,
      "loss": 0.8887,
      "step": 543
    },
    {
      "epoch": 0.2197868794505328,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0002873058252427184,
      "loss": 0.9723,
      "step": 544
    },
    {
      "epoch": 0.22019089944952275,
      "grad_norm": 0.546875,
      "learning_rate": 0.00028728155339805824,
      "loss": 0.8342,
      "step": 545
    },
    {
      "epoch": 0.2205949194485127,
      "grad_norm": 0.68359375,
      "learning_rate": 0.000287257281553398,
      "loss": 0.9256,
      "step": 546
    },
    {
      "epoch": 0.22099893944750265,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00028723300970873785,
      "loss": 0.924,
      "step": 547
    },
    {
      "epoch": 0.2214029594464926,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002872087378640776,
      "loss": 0.9493,
      "step": 548
    },
    {
      "epoch": 0.22180697944548255,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00028718446601941745,
      "loss": 0.7858,
      "step": 549
    },
    {
      "epoch": 0.2222109994444725,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002871601941747573,
      "loss": 0.9084,
      "step": 550
    },
    {
      "epoch": 0.22261501944346246,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00028713592233009706,
      "loss": 0.887,
      "step": 551
    },
    {
      "epoch": 0.2230190394424524,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0002871116504854369,
      "loss": 1.0842,
      "step": 552
    },
    {
      "epoch": 0.22342305944144236,
      "grad_norm": 0.734375,
      "learning_rate": 0.00028708737864077666,
      "loss": 0.9642,
      "step": 553
    },
    {
      "epoch": 0.2238270794404323,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002870631067961165,
      "loss": 0.9266,
      "step": 554
    },
    {
      "epoch": 0.22423109943942224,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002870388349514563,
      "loss": 0.8952,
      "step": 555
    },
    {
      "epoch": 0.2246351194384122,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002870145631067961,
      "loss": 0.8276,
      "step": 556
    },
    {
      "epoch": 0.22503913943740214,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00028699029126213587,
      "loss": 0.8826,
      "step": 557
    },
    {
      "epoch": 0.2254431594363921,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0002869660194174757,
      "loss": 0.8993,
      "step": 558
    },
    {
      "epoch": 0.22584717943538204,
      "grad_norm": 0.640625,
      "learning_rate": 0.00028694174757281553,
      "loss": 0.9939,
      "step": 559
    },
    {
      "epoch": 0.226251199434372,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002869174757281553,
      "loss": 0.8858,
      "step": 560
    },
    {
      "epoch": 0.22665521943336195,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00028689320388349513,
      "loss": 1.034,
      "step": 561
    },
    {
      "epoch": 0.2270592394323519,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0002868689320388349,
      "loss": 0.9795,
      "step": 562
    },
    {
      "epoch": 0.22746325943134185,
      "grad_norm": 0.765625,
      "learning_rate": 0.00028684466019417474,
      "loss": 1.0035,
      "step": 563
    },
    {
      "epoch": 0.2278672794303318,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0002868203883495145,
      "loss": 0.9007,
      "step": 564
    },
    {
      "epoch": 0.22827129942932176,
      "grad_norm": 1.125,
      "learning_rate": 0.00028679611650485434,
      "loss": 0.9216,
      "step": 565
    },
    {
      "epoch": 0.2286753194283117,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002867718446601942,
      "loss": 0.8631,
      "step": 566
    },
    {
      "epoch": 0.22907933942730166,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00028674757281553395,
      "loss": 1.0407,
      "step": 567
    },
    {
      "epoch": 0.2294833594262916,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0002867233009708737,
      "loss": 0.8925,
      "step": 568
    },
    {
      "epoch": 0.22988737942528156,
      "grad_norm": 1.390625,
      "learning_rate": 0.00028669902912621355,
      "loss": 1.1478,
      "step": 569
    },
    {
      "epoch": 0.23029139942427151,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0002866747572815534,
      "loss": 0.9599,
      "step": 570
    },
    {
      "epoch": 0.23069541942326144,
      "grad_norm": 0.796875,
      "learning_rate": 0.00028665048543689316,
      "loss": 1.0164,
      "step": 571
    },
    {
      "epoch": 0.2310994394222514,
      "grad_norm": 0.69921875,
      "learning_rate": 0.000286626213592233,
      "loss": 0.8927,
      "step": 572
    },
    {
      "epoch": 0.23150345942124134,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00028660194174757276,
      "loss": 0.9198,
      "step": 573
    },
    {
      "epoch": 0.2319074794202313,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0002865776699029126,
      "loss": 1.0202,
      "step": 574
    },
    {
      "epoch": 0.23231149941922125,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0002865533980582524,
      "loss": 0.9704,
      "step": 575
    },
    {
      "epoch": 0.2327155194182112,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0002865291262135922,
      "loss": 1.1209,
      "step": 576
    },
    {
      "epoch": 0.23311953941720115,
      "grad_norm": 0.70703125,
      "learning_rate": 0.000286504854368932,
      "loss": 1.0094,
      "step": 577
    },
    {
      "epoch": 0.2335235594161911,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002864805825242718,
      "loss": 0.8508,
      "step": 578
    },
    {
      "epoch": 0.23392757941518105,
      "grad_norm": 0.734375,
      "learning_rate": 0.00028645631067961163,
      "loss": 1.1793,
      "step": 579
    },
    {
      "epoch": 0.234331599414171,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00028643203883495146,
      "loss": 0.9887,
      "step": 580
    },
    {
      "epoch": 0.23473561941316096,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00028640776699029124,
      "loss": 0.9601,
      "step": 581
    },
    {
      "epoch": 0.2351396394121509,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00028638349514563107,
      "loss": 1.0661,
      "step": 582
    },
    {
      "epoch": 0.23554365941114086,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00028635922330097084,
      "loss": 0.8554,
      "step": 583
    },
    {
      "epoch": 0.2359476794101308,
      "grad_norm": 0.859375,
      "learning_rate": 0.00028633495145631067,
      "loss": 0.8983,
      "step": 584
    },
    {
      "epoch": 0.23635169940912076,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002863106796116505,
      "loss": 0.9391,
      "step": 585
    },
    {
      "epoch": 0.23675571940811072,
      "grad_norm": 2.328125,
      "learning_rate": 0.0002862864077669903,
      "loss": 0.9954,
      "step": 586
    },
    {
      "epoch": 0.23715973940710064,
      "grad_norm": 0.578125,
      "learning_rate": 0.00028626213592233005,
      "loss": 0.8463,
      "step": 587
    },
    {
      "epoch": 0.2375637594060906,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002862378640776699,
      "loss": 0.8266,
      "step": 588
    },
    {
      "epoch": 0.23796777940508054,
      "grad_norm": 0.75390625,
      "learning_rate": 0.00028621359223300965,
      "loss": 1.1127,
      "step": 589
    },
    {
      "epoch": 0.2383717994040705,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0002861893203883495,
      "loss": 1.0256,
      "step": 590
    },
    {
      "epoch": 0.23877581940306045,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002861650485436893,
      "loss": 0.834,
      "step": 591
    },
    {
      "epoch": 0.2391798394020504,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002861407766990291,
      "loss": 0.8038,
      "step": 592
    },
    {
      "epoch": 0.23958385940104035,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002861165048543689,
      "loss": 0.9478,
      "step": 593
    },
    {
      "epoch": 0.2399878794000303,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0002860922330097087,
      "loss": 1.0882,
      "step": 594
    },
    {
      "epoch": 0.24039189939902025,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002860679611650485,
      "loss": 0.9208,
      "step": 595
    },
    {
      "epoch": 0.2407959193980102,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00028604368932038835,
      "loss": 0.9013,
      "step": 596
    },
    {
      "epoch": 0.24119993939700016,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00028601941747572813,
      "loss": 0.8794,
      "step": 597
    },
    {
      "epoch": 0.2416039593959901,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002859951456310679,
      "loss": 0.9299,
      "step": 598
    },
    {
      "epoch": 0.24200797939498006,
      "grad_norm": 0.78125,
      "learning_rate": 0.00028597087378640773,
      "loss": 1.034,
      "step": 599
    },
    {
      "epoch": 0.24241199939397,
      "grad_norm": 0.640625,
      "learning_rate": 0.00028594660194174756,
      "loss": 0.9038,
      "step": 600
    },
    {
      "epoch": 0.24281601939295996,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00028592233009708734,
      "loss": 0.9749,
      "step": 601
    },
    {
      "epoch": 0.2432200393919499,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00028589805825242717,
      "loss": 0.8349,
      "step": 602
    },
    {
      "epoch": 0.24362405939093984,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00028587378640776694,
      "loss": 0.9917,
      "step": 603
    },
    {
      "epoch": 0.2440280793899298,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00028584951456310677,
      "loss": 0.9405,
      "step": 604
    },
    {
      "epoch": 0.24443209938891974,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0002858252427184466,
      "loss": 0.8991,
      "step": 605
    },
    {
      "epoch": 0.2448361193879097,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0002858009708737864,
      "loss": 0.9084,
      "step": 606
    },
    {
      "epoch": 0.24524013938689965,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002857766990291262,
      "loss": 0.9519,
      "step": 607
    },
    {
      "epoch": 0.2456441593858896,
      "grad_norm": 0.640625,
      "learning_rate": 0.000285752427184466,
      "loss": 1.0394,
      "step": 608
    },
    {
      "epoch": 0.24604817938487955,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002857281553398058,
      "loss": 0.822,
      "step": 609
    },
    {
      "epoch": 0.2464521993838695,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00028570388349514564,
      "loss": 0.8921,
      "step": 610
    },
    {
      "epoch": 0.24685621938285945,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002856796116504854,
      "loss": 1.0395,
      "step": 611
    },
    {
      "epoch": 0.2472602393818494,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00028565533980582524,
      "loss": 1.0082,
      "step": 612
    },
    {
      "epoch": 0.24766425938083936,
      "grad_norm": 0.546875,
      "learning_rate": 0.000285631067961165,
      "loss": 0.8228,
      "step": 613
    },
    {
      "epoch": 0.2480682793798293,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002856067961165048,
      "loss": 0.9422,
      "step": 614
    },
    {
      "epoch": 0.24847229937881926,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002855825242718446,
      "loss": 0.8533,
      "step": 615
    },
    {
      "epoch": 0.2488763193778092,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00028555825242718445,
      "loss": 0.9887,
      "step": 616
    },
    {
      "epoch": 0.24928033937679916,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00028553398058252423,
      "loss": 0.9982,
      "step": 617
    },
    {
      "epoch": 0.2496843593757891,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00028550970873786406,
      "loss": 0.9553,
      "step": 618
    },
    {
      "epoch": 0.25008837937477907,
      "grad_norm": 0.71875,
      "learning_rate": 0.00028548543689320383,
      "loss": 1.0163,
      "step": 619
    },
    {
      "epoch": 0.250492399373769,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00028546116504854366,
      "loss": 0.8816,
      "step": 620
    },
    {
      "epoch": 0.25089641937275897,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002854368932038835,
      "loss": 0.8976,
      "step": 621
    },
    {
      "epoch": 0.2513004393717489,
      "grad_norm": 0.671875,
      "learning_rate": 0.00028541262135922327,
      "loss": 1.0007,
      "step": 622
    },
    {
      "epoch": 0.2517044593707389,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002853883495145631,
      "loss": 0.8655,
      "step": 623
    },
    {
      "epoch": 0.2521084793697288,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00028536407766990287,
      "loss": 0.8199,
      "step": 624
    },
    {
      "epoch": 0.2525124993687188,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0002853398058252427,
      "loss": 0.9979,
      "step": 625
    },
    {
      "epoch": 0.2529165193677087,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00028531553398058253,
      "loss": 0.8656,
      "step": 626
    },
    {
      "epoch": 0.2533205393666986,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002852912621359223,
      "loss": 0.9553,
      "step": 627
    },
    {
      "epoch": 0.2537245593656886,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002852669902912621,
      "loss": 0.9857,
      "step": 628
    },
    {
      "epoch": 0.25412857936467853,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002852427184466019,
      "loss": 1.0002,
      "step": 629
    },
    {
      "epoch": 0.2545325993636685,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00028521844660194174,
      "loss": 1.0198,
      "step": 630
    },
    {
      "epoch": 0.25493661936265843,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002851941747572815,
      "loss": 0.9464,
      "step": 631
    },
    {
      "epoch": 0.2553406393616484,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00028516990291262135,
      "loss": 0.9267,
      "step": 632
    },
    {
      "epoch": 0.25574465936063834,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0002851456310679611,
      "loss": 1.0059,
      "step": 633
    },
    {
      "epoch": 0.2561486793596283,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00028512135922330095,
      "loss": 0.8549,
      "step": 634
    },
    {
      "epoch": 0.25655269935861824,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002850970873786408,
      "loss": 0.9661,
      "step": 635
    },
    {
      "epoch": 0.2569567193576082,
      "grad_norm": 0.625,
      "learning_rate": 0.00028507281553398055,
      "loss": 0.8712,
      "step": 636
    },
    {
      "epoch": 0.25736073935659814,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002850485436893204,
      "loss": 0.8885,
      "step": 637
    },
    {
      "epoch": 0.2577647593555881,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00028502427184466016,
      "loss": 0.8802,
      "step": 638
    },
    {
      "epoch": 0.25816877935457805,
      "grad_norm": 0.67578125,
      "learning_rate": 0.000285,
      "loss": 0.8805,
      "step": 639
    },
    {
      "epoch": 0.25857279935356803,
      "grad_norm": 0.59375,
      "learning_rate": 0.00028497572815533976,
      "loss": 0.8873,
      "step": 640
    },
    {
      "epoch": 0.25897681935255795,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0002849514563106796,
      "loss": 1.1061,
      "step": 641
    },
    {
      "epoch": 0.2593808393515479,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00028492718446601937,
      "loss": 0.8934,
      "step": 642
    },
    {
      "epoch": 0.25978485935053786,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002849029126213592,
      "loss": 0.9627,
      "step": 643
    },
    {
      "epoch": 0.2601888793495278,
      "grad_norm": 0.57421875,
      "learning_rate": 0.000284878640776699,
      "loss": 0.8594,
      "step": 644
    },
    {
      "epoch": 0.26059289934851776,
      "grad_norm": 0.796875,
      "learning_rate": 0.0002848543689320388,
      "loss": 1.0497,
      "step": 645
    },
    {
      "epoch": 0.2609969193475077,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00028483009708737863,
      "loss": 0.8962,
      "step": 646
    },
    {
      "epoch": 0.26140093934649766,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002848058252427184,
      "loss": 0.9408,
      "step": 647
    },
    {
      "epoch": 0.2618049593454876,
      "grad_norm": 0.578125,
      "learning_rate": 0.00028478155339805824,
      "loss": 0.9544,
      "step": 648
    },
    {
      "epoch": 0.26220897934447757,
      "grad_norm": 0.66796875,
      "learning_rate": 0.000284757281553398,
      "loss": 0.9871,
      "step": 649
    },
    {
      "epoch": 0.2626129993434675,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00028473300970873784,
      "loss": 0.893,
      "step": 650
    },
    {
      "epoch": 0.26301701934245747,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00028470873786407767,
      "loss": 1.0196,
      "step": 651
    },
    {
      "epoch": 0.2634210393414474,
      "grad_norm": 0.59375,
      "learning_rate": 0.00028468446601941745,
      "loss": 0.8983,
      "step": 652
    },
    {
      "epoch": 0.2638250593404374,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002846601941747573,
      "loss": 0.921,
      "step": 653
    },
    {
      "epoch": 0.2642290793394273,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00028463592233009705,
      "loss": 1.0104,
      "step": 654
    },
    {
      "epoch": 0.2646330993384173,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002846116504854369,
      "loss": 0.9128,
      "step": 655
    },
    {
      "epoch": 0.2650371193374072,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002845873786407767,
      "loss": 0.8141,
      "step": 656
    },
    {
      "epoch": 0.2654411393363971,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002845631067961165,
      "loss": 0.9438,
      "step": 657
    },
    {
      "epoch": 0.2658451593353871,
      "grad_norm": 0.640625,
      "learning_rate": 0.00028453883495145626,
      "loss": 0.9011,
      "step": 658
    },
    {
      "epoch": 0.26624917933437703,
      "grad_norm": 4.71875,
      "learning_rate": 0.0002845145631067961,
      "loss": 1.2513,
      "step": 659
    },
    {
      "epoch": 0.266653199333367,
      "grad_norm": 0.84375,
      "learning_rate": 0.0002844902912621359,
      "loss": 1.1288,
      "step": 660
    },
    {
      "epoch": 0.26705721933235693,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002844660194174757,
      "loss": 1.1012,
      "step": 661
    },
    {
      "epoch": 0.2674612393313469,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002844417475728155,
      "loss": 0.9636,
      "step": 662
    },
    {
      "epoch": 0.26786525933033684,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002844174757281553,
      "loss": 0.9212,
      "step": 663
    },
    {
      "epoch": 0.2682692793293268,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00028439320388349513,
      "loss": 0.9566,
      "step": 664
    },
    {
      "epoch": 0.26867329932831674,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002843689320388349,
      "loss": 1.0172,
      "step": 665
    },
    {
      "epoch": 0.2690773193273067,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00028434466019417473,
      "loss": 0.9269,
      "step": 666
    },
    {
      "epoch": 0.26948133932629664,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00028432038834951456,
      "loss": 0.8695,
      "step": 667
    },
    {
      "epoch": 0.2698853593252866,
      "grad_norm": 0.7421875,
      "learning_rate": 0.00028429611650485434,
      "loss": 0.9769,
      "step": 668
    },
    {
      "epoch": 0.27028937932427655,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002842718446601941,
      "loss": 0.9047,
      "step": 669
    },
    {
      "epoch": 0.2706933993232665,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00028424757281553394,
      "loss": 0.9375,
      "step": 670
    },
    {
      "epoch": 0.27109741932225645,
      "grad_norm": 0.609375,
      "learning_rate": 0.00028422330097087377,
      "loss": 0.9032,
      "step": 671
    },
    {
      "epoch": 0.27150143932124643,
      "grad_norm": 0.671875,
      "learning_rate": 0.00028419902912621355,
      "loss": 0.9476,
      "step": 672
    },
    {
      "epoch": 0.27190545932023635,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002841747572815534,
      "loss": 0.9191,
      "step": 673
    },
    {
      "epoch": 0.2723094793192263,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00028415048543689315,
      "loss": 1.0055,
      "step": 674
    },
    {
      "epoch": 0.27271349931821626,
      "grad_norm": 0.53515625,
      "learning_rate": 0.000284126213592233,
      "loss": 0.8339,
      "step": 675
    },
    {
      "epoch": 0.2731175193172062,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002841019417475728,
      "loss": 0.9243,
      "step": 676
    },
    {
      "epoch": 0.27352153931619616,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002840776699029126,
      "loss": 0.842,
      "step": 677
    },
    {
      "epoch": 0.2739255593151861,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002840533980582524,
      "loss": 0.8922,
      "step": 678
    },
    {
      "epoch": 0.27432957931417606,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002840291262135922,
      "loss": 0.9799,
      "step": 679
    },
    {
      "epoch": 0.274733599313166,
      "grad_norm": 0.62109375,
      "learning_rate": 0.000284004854368932,
      "loss": 0.8862,
      "step": 680
    },
    {
      "epoch": 0.27513761931215597,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00028398058252427185,
      "loss": 0.8683,
      "step": 681
    },
    {
      "epoch": 0.2755416393111459,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002839563106796116,
      "loss": 0.9163,
      "step": 682
    },
    {
      "epoch": 0.27594565931013587,
      "grad_norm": 0.7578125,
      "learning_rate": 0.00028393203883495145,
      "loss": 1.1206,
      "step": 683
    },
    {
      "epoch": 0.2763496793091258,
      "grad_norm": 0.59375,
      "learning_rate": 0.00028390776699029123,
      "loss": 0.9849,
      "step": 684
    },
    {
      "epoch": 0.2767536993081158,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00028388349514563106,
      "loss": 0.8585,
      "step": 685
    },
    {
      "epoch": 0.2771577193071057,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002838592233009709,
      "loss": 0.9802,
      "step": 686
    },
    {
      "epoch": 0.2775617393060957,
      "grad_norm": 0.640625,
      "learning_rate": 0.00028383495145631066,
      "loss": 1.0202,
      "step": 687
    },
    {
      "epoch": 0.2779657593050856,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00028381067961165044,
      "loss": 1.012,
      "step": 688
    },
    {
      "epoch": 0.2783697793040755,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00028378640776699027,
      "loss": 0.9712,
      "step": 689
    },
    {
      "epoch": 0.2787737993030655,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00028376213592233004,
      "loss": 0.9298,
      "step": 690
    },
    {
      "epoch": 0.27917781930205543,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002837378640776699,
      "loss": 0.9768,
      "step": 691
    },
    {
      "epoch": 0.2795818393010454,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002837135922330097,
      "loss": 0.903,
      "step": 692
    },
    {
      "epoch": 0.27998585930003533,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0002836893203883495,
      "loss": 0.9531,
      "step": 693
    },
    {
      "epoch": 0.2803898792990253,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002836650485436893,
      "loss": 0.8968,
      "step": 694
    },
    {
      "epoch": 0.28079389929801524,
      "grad_norm": 3.109375,
      "learning_rate": 0.0002836407766990291,
      "loss": 0.8792,
      "step": 695
    },
    {
      "epoch": 0.2811979192970052,
      "grad_norm": 0.84375,
      "learning_rate": 0.0002836165048543689,
      "loss": 1.1256,
      "step": 696
    },
    {
      "epoch": 0.28160193929599514,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00028359223300970874,
      "loss": 0.7954,
      "step": 697
    },
    {
      "epoch": 0.2820059592949851,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002835679611650485,
      "loss": 0.9192,
      "step": 698
    },
    {
      "epoch": 0.28240997929397504,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002835436893203883,
      "loss": 0.9071,
      "step": 699
    },
    {
      "epoch": 0.282813999292965,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002835194174757281,
      "loss": 0.9338,
      "step": 700
    },
    {
      "epoch": 0.28321801929195495,
      "grad_norm": 0.640625,
      "learning_rate": 0.00028349514563106795,
      "loss": 0.8559,
      "step": 701
    },
    {
      "epoch": 0.2836220392909449,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002834708737864077,
      "loss": 0.8797,
      "step": 702
    },
    {
      "epoch": 0.28402605928993485,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00028344660194174756,
      "loss": 0.9025,
      "step": 703
    },
    {
      "epoch": 0.2844300792889248,
      "grad_norm": 0.6875,
      "learning_rate": 0.00028342233009708733,
      "loss": 0.9541,
      "step": 704
    },
    {
      "epoch": 0.28483409928791475,
      "grad_norm": 0.609375,
      "learning_rate": 0.00028339805825242716,
      "loss": 0.8582,
      "step": 705
    },
    {
      "epoch": 0.2852381192869047,
      "grad_norm": 0.71875,
      "learning_rate": 0.000283373786407767,
      "loss": 1.0486,
      "step": 706
    },
    {
      "epoch": 0.28564213928589466,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00028334951456310676,
      "loss": 0.9017,
      "step": 707
    },
    {
      "epoch": 0.2860461592848846,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002833252427184466,
      "loss": 0.9046,
      "step": 708
    },
    {
      "epoch": 0.28645017928387456,
      "grad_norm": 0.703125,
      "learning_rate": 0.00028330097087378637,
      "loss": 0.96,
      "step": 709
    },
    {
      "epoch": 0.2868541992828645,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002832766990291262,
      "loss": 0.826,
      "step": 710
    },
    {
      "epoch": 0.28725821928185447,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00028325242718446603,
      "loss": 0.8317,
      "step": 711
    },
    {
      "epoch": 0.2876622392808444,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0002832281553398058,
      "loss": 1.112,
      "step": 712
    },
    {
      "epoch": 0.28806625927983437,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00028320388349514563,
      "loss": 0.8887,
      "step": 713
    },
    {
      "epoch": 0.2884702792788243,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002831796116504854,
      "loss": 0.8881,
      "step": 714
    },
    {
      "epoch": 0.2888742992778143,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002831553398058252,
      "loss": 0.8682,
      "step": 715
    },
    {
      "epoch": 0.2892783192768042,
      "grad_norm": 0.60546875,
      "learning_rate": 0.000283131067961165,
      "loss": 0.9852,
      "step": 716
    },
    {
      "epoch": 0.2896823392757942,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00028310679611650484,
      "loss": 0.8942,
      "step": 717
    },
    {
      "epoch": 0.2900863592747841,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002830825242718446,
      "loss": 0.8331,
      "step": 718
    },
    {
      "epoch": 0.290490379273774,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00028305825242718445,
      "loss": 0.9571,
      "step": 719
    },
    {
      "epoch": 0.290894399272764,
      "grad_norm": 7.125,
      "learning_rate": 0.0002830339805825242,
      "loss": 1.1611,
      "step": 720
    },
    {
      "epoch": 0.29129841927175393,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00028300970873786405,
      "loss": 1.0906,
      "step": 721
    },
    {
      "epoch": 0.2917024392707439,
      "grad_norm": 0.78125,
      "learning_rate": 0.0002829854368932039,
      "loss": 0.7732,
      "step": 722
    },
    {
      "epoch": 0.29210645926973383,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00028296116504854366,
      "loss": 0.8838,
      "step": 723
    },
    {
      "epoch": 0.2925104792687238,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0002829368932038835,
      "loss": 1.0307,
      "step": 724
    },
    {
      "epoch": 0.29291449926771373,
      "grad_norm": 0.734375,
      "learning_rate": 0.00028291262135922326,
      "loss": 0.9486,
      "step": 725
    },
    {
      "epoch": 0.2933185192667037,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002828883495145631,
      "loss": 0.8739,
      "step": 726
    },
    {
      "epoch": 0.29372253926569364,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002828640776699029,
      "loss": 0.7921,
      "step": 727
    },
    {
      "epoch": 0.2941265592646836,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002828398058252427,
      "loss": 1.1104,
      "step": 728
    },
    {
      "epoch": 0.29453057926367354,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00028281553398058247,
      "loss": 0.8588,
      "step": 729
    },
    {
      "epoch": 0.2949345992626635,
      "grad_norm": 2.171875,
      "learning_rate": 0.0002827912621359223,
      "loss": 1.0735,
      "step": 730
    },
    {
      "epoch": 0.29533861926165345,
      "grad_norm": 2.265625,
      "learning_rate": 0.00028276699029126213,
      "loss": 0.9293,
      "step": 731
    },
    {
      "epoch": 0.2957426392606434,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002827427184466019,
      "loss": 1.1045,
      "step": 732
    },
    {
      "epoch": 0.29614665925963335,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00028271844660194173,
      "loss": 0.974,
      "step": 733
    },
    {
      "epoch": 0.29655067925862333,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0002826941747572815,
      "loss": 0.8526,
      "step": 734
    },
    {
      "epoch": 0.29695469925761325,
      "grad_norm": 3.421875,
      "learning_rate": 0.00028266990291262134,
      "loss": 0.9843,
      "step": 735
    },
    {
      "epoch": 0.2973587192566032,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00028264563106796117,
      "loss": 0.9592,
      "step": 736
    },
    {
      "epoch": 0.29776273925559316,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00028262135922330094,
      "loss": 0.8841,
      "step": 737
    },
    {
      "epoch": 0.2981667592545831,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00028259708737864077,
      "loss": 1.0498,
      "step": 738
    },
    {
      "epoch": 0.29857077925357306,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00028257281553398055,
      "loss": 0.9624,
      "step": 739
    },
    {
      "epoch": 0.298974799252563,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002825485436893204,
      "loss": 0.9093,
      "step": 740
    },
    {
      "epoch": 0.29937881925155296,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00028252427184466015,
      "loss": 0.8723,
      "step": 741
    },
    {
      "epoch": 0.2997828392505429,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002825,
      "loss": 0.7886,
      "step": 742
    },
    {
      "epoch": 0.30018685924953287,
      "grad_norm": 0.65625,
      "learning_rate": 0.00028247572815533976,
      "loss": 0.9585,
      "step": 743
    },
    {
      "epoch": 0.3005908792485228,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002824514563106796,
      "loss": 0.9489,
      "step": 744
    },
    {
      "epoch": 0.30099489924751277,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00028242718446601936,
      "loss": 0.9185,
      "step": 745
    },
    {
      "epoch": 0.3013989192465027,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0002824029126213592,
      "loss": 0.9173,
      "step": 746
    },
    {
      "epoch": 0.3018029392454927,
      "grad_norm": 0.70703125,
      "learning_rate": 0.000282378640776699,
      "loss": 0.9368,
      "step": 747
    },
    {
      "epoch": 0.3022069592444826,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002823543689320388,
      "loss": 1.0581,
      "step": 748
    },
    {
      "epoch": 0.3026109792434726,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002823300970873786,
      "loss": 1.091,
      "step": 749
    },
    {
      "epoch": 0.3030149992424625,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0002823058252427184,
      "loss": 0.8759,
      "step": 750
    },
    {
      "epoch": 0.3034190192414524,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00028228155339805823,
      "loss": 0.8685,
      "step": 751
    },
    {
      "epoch": 0.3038230392404424,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028225728155339806,
      "loss": 0.9998,
      "step": 752
    },
    {
      "epoch": 0.30422705923943233,
      "grad_norm": 0.609375,
      "learning_rate": 0.00028223300970873783,
      "loss": 0.9506,
      "step": 753
    },
    {
      "epoch": 0.3046310792384223,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00028220873786407766,
      "loss": 0.9506,
      "step": 754
    },
    {
      "epoch": 0.30503509923741223,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00028218446601941744,
      "loss": 0.8966,
      "step": 755
    },
    {
      "epoch": 0.3054391192364022,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00028216019417475727,
      "loss": 0.8222,
      "step": 756
    },
    {
      "epoch": 0.30584313923539214,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0002821359223300971,
      "loss": 1.0288,
      "step": 757
    },
    {
      "epoch": 0.3062471592343821,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002821116504854369,
      "loss": 0.7993,
      "step": 758
    },
    {
      "epoch": 0.30665117923337204,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00028208737864077665,
      "loss": 1.0813,
      "step": 759
    },
    {
      "epoch": 0.307055199232362,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002820631067961165,
      "loss": 0.9298,
      "step": 760
    },
    {
      "epoch": 0.30745921923135194,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002820388349514563,
      "loss": 0.796,
      "step": 761
    },
    {
      "epoch": 0.3078632392303419,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0002820145631067961,
      "loss": 1.0613,
      "step": 762
    },
    {
      "epoch": 0.30826725922933185,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0002819902912621359,
      "loss": 0.9907,
      "step": 763
    },
    {
      "epoch": 0.3086712792283218,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002819660194174757,
      "loss": 0.8779,
      "step": 764
    },
    {
      "epoch": 0.30907529922731175,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0002819417475728155,
      "loss": 0.8005,
      "step": 765
    },
    {
      "epoch": 0.3094793192263017,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002819174757281553,
      "loss": 1.0027,
      "step": 766
    },
    {
      "epoch": 0.30988333922529165,
      "grad_norm": 3.390625,
      "learning_rate": 0.0002818932038834951,
      "loss": 1.2817,
      "step": 767
    },
    {
      "epoch": 0.3102873592242816,
      "grad_norm": 0.78125,
      "learning_rate": 0.00028186893203883495,
      "loss": 1.1035,
      "step": 768
    },
    {
      "epoch": 0.31069137922327156,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002818446601941747,
      "loss": 0.9828,
      "step": 769
    },
    {
      "epoch": 0.3110953992222615,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002818203883495145,
      "loss": 0.9242,
      "step": 770
    },
    {
      "epoch": 0.31149941922125146,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00028179611650485433,
      "loss": 0.7967,
      "step": 771
    },
    {
      "epoch": 0.3119034392202414,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00028177184466019416,
      "loss": 0.9886,
      "step": 772
    },
    {
      "epoch": 0.31230745921923136,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028174757281553394,
      "loss": 0.914,
      "step": 773
    },
    {
      "epoch": 0.3127114792182213,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00028172330097087377,
      "loss": 0.9278,
      "step": 774
    },
    {
      "epoch": 0.31311549921721127,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00028169902912621354,
      "loss": 0.867,
      "step": 775
    },
    {
      "epoch": 0.3135195192162012,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00028167475728155337,
      "loss": 0.8429,
      "step": 776
    },
    {
      "epoch": 0.31392353921519117,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002816504854368932,
      "loss": 1.0028,
      "step": 777
    },
    {
      "epoch": 0.3143275592141811,
      "grad_norm": 0.6796875,
      "learning_rate": 0.000281626213592233,
      "loss": 0.9794,
      "step": 778
    },
    {
      "epoch": 0.3147315792131711,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002816019417475728,
      "loss": 0.8115,
      "step": 779
    },
    {
      "epoch": 0.315135599212161,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002815776699029126,
      "loss": 0.783,
      "step": 780
    },
    {
      "epoch": 0.315539619211151,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002815533980582524,
      "loss": 0.7876,
      "step": 781
    },
    {
      "epoch": 0.3159436392101409,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00028152912621359224,
      "loss": 0.792,
      "step": 782
    },
    {
      "epoch": 0.3163476592091308,
      "grad_norm": 0.6953125,
      "learning_rate": 0.000281504854368932,
      "loss": 0.8519,
      "step": 783
    },
    {
      "epoch": 0.3167516792081208,
      "grad_norm": 0.640625,
      "learning_rate": 0.00028148058252427184,
      "loss": 0.9951,
      "step": 784
    },
    {
      "epoch": 0.31715569920711073,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002814563106796116,
      "loss": 0.91,
      "step": 785
    },
    {
      "epoch": 0.3175597192061007,
      "grad_norm": 0.578125,
      "learning_rate": 0.00028143203883495145,
      "loss": 0.9021,
      "step": 786
    },
    {
      "epoch": 0.31796373920509063,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002814077669902913,
      "loss": 0.7973,
      "step": 787
    },
    {
      "epoch": 0.3183677592040806,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00028138349514563105,
      "loss": 0.9338,
      "step": 788
    },
    {
      "epoch": 0.31877177920307054,
      "grad_norm": 0.78125,
      "learning_rate": 0.00028135922330097083,
      "loss": 0.9241,
      "step": 789
    },
    {
      "epoch": 0.3191757992020605,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00028133495145631066,
      "loss": 0.8217,
      "step": 790
    },
    {
      "epoch": 0.31957981920105044,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00028131067961165043,
      "loss": 0.8936,
      "step": 791
    },
    {
      "epoch": 0.3199838392000404,
      "grad_norm": 0.8125,
      "learning_rate": 0.00028128640776699026,
      "loss": 0.9677,
      "step": 792
    },
    {
      "epoch": 0.32038785919903034,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002812621359223301,
      "loss": 0.9456,
      "step": 793
    },
    {
      "epoch": 0.3207918791980203,
      "grad_norm": 0.609375,
      "learning_rate": 0.00028123786407766987,
      "loss": 0.9369,
      "step": 794
    },
    {
      "epoch": 0.32119589919701025,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002812135922330097,
      "loss": 0.8844,
      "step": 795
    },
    {
      "epoch": 0.32159991919600023,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00028118932038834947,
      "loss": 0.8333,
      "step": 796
    },
    {
      "epoch": 0.32200393919499015,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002811650485436893,
      "loss": 0.908,
      "step": 797
    },
    {
      "epoch": 0.3224079591939801,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00028114077669902913,
      "loss": 0.8693,
      "step": 798
    },
    {
      "epoch": 0.32281197919297006,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0002811165048543689,
      "loss": 0.9163,
      "step": 799
    },
    {
      "epoch": 0.32321599919196,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002810922330097087,
      "loss": 0.9512,
      "step": 800
    },
    {
      "epoch": 0.32362001919094996,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002810679611650485,
      "loss": 0.8617,
      "step": 801
    },
    {
      "epoch": 0.3240240391899399,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00028104368932038834,
      "loss": 0.8462,
      "step": 802
    },
    {
      "epoch": 0.32442805918892986,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002810194174757281,
      "loss": 0.9403,
      "step": 803
    },
    {
      "epoch": 0.3248320791879198,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00028099514563106794,
      "loss": 0.9139,
      "step": 804
    },
    {
      "epoch": 0.32523609918690977,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002809708737864077,
      "loss": 0.8284,
      "step": 805
    },
    {
      "epoch": 0.3256401191858997,
      "grad_norm": 0.78125,
      "learning_rate": 0.00028094660194174755,
      "loss": 1.1109,
      "step": 806
    },
    {
      "epoch": 0.32604413918488967,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0002809223300970874,
      "loss": 0.9832,
      "step": 807
    },
    {
      "epoch": 0.3264481591838796,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00028089805825242715,
      "loss": 0.8983,
      "step": 808
    },
    {
      "epoch": 0.3268521791828696,
      "grad_norm": 0.68359375,
      "learning_rate": 0.000280873786407767,
      "loss": 1.0228,
      "step": 809
    },
    {
      "epoch": 0.3272561991818595,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00028084951456310676,
      "loss": 0.7923,
      "step": 810
    },
    {
      "epoch": 0.3276602191808495,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002808252427184466,
      "loss": 0.9158,
      "step": 811
    },
    {
      "epoch": 0.3280642391798394,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002808009708737864,
      "loss": 0.8969,
      "step": 812
    },
    {
      "epoch": 0.3284682591788293,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0002807766990291262,
      "loss": 1.0301,
      "step": 813
    },
    {
      "epoch": 0.3288722791778193,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000280752427184466,
      "loss": 0.9933,
      "step": 814
    },
    {
      "epoch": 0.32927629917680923,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002807281553398058,
      "loss": 0.886,
      "step": 815
    },
    {
      "epoch": 0.3296803191757992,
      "grad_norm": 0.546875,
      "learning_rate": 0.00028070388349514557,
      "loss": 0.8494,
      "step": 816
    },
    {
      "epoch": 0.33008433917478913,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002806796116504854,
      "loss": 0.9817,
      "step": 817
    },
    {
      "epoch": 0.3304883591737791,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00028065533980582523,
      "loss": 0.9445,
      "step": 818
    },
    {
      "epoch": 0.33089237917276904,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000280631067961165,
      "loss": 0.9852,
      "step": 819
    },
    {
      "epoch": 0.331296399171759,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00028060679611650484,
      "loss": 0.9259,
      "step": 820
    },
    {
      "epoch": 0.33170041917074894,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002805825242718446,
      "loss": 0.8589,
      "step": 821
    },
    {
      "epoch": 0.3321044391697389,
      "grad_norm": 0.6875,
      "learning_rate": 0.00028055825242718444,
      "loss": 0.9431,
      "step": 822
    },
    {
      "epoch": 0.33250845916872884,
      "grad_norm": 0.78125,
      "learning_rate": 0.00028053398058252427,
      "loss": 0.9913,
      "step": 823
    },
    {
      "epoch": 0.3329124791677188,
      "grad_norm": 0.796875,
      "learning_rate": 0.00028050970873786405,
      "loss": 0.9774,
      "step": 824
    },
    {
      "epoch": 0.33331649916670875,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0002804854368932039,
      "loss": 0.9538,
      "step": 825
    },
    {
      "epoch": 0.3337205191656987,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00028046116504854365,
      "loss": 0.9812,
      "step": 826
    },
    {
      "epoch": 0.33412453916468865,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002804368932038835,
      "loss": 0.9436,
      "step": 827
    },
    {
      "epoch": 0.33452855916367863,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002804126213592233,
      "loss": 0.9036,
      "step": 828
    },
    {
      "epoch": 0.33493257916266855,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002803883495145631,
      "loss": 0.9794,
      "step": 829
    },
    {
      "epoch": 0.3353365991616585,
      "grad_norm": 0.7421875,
      "learning_rate": 0.00028036407766990286,
      "loss": 0.8456,
      "step": 830
    },
    {
      "epoch": 0.33574061916064846,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002803398058252427,
      "loss": 0.8338,
      "step": 831
    },
    {
      "epoch": 0.3361446391596384,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0002803155339805825,
      "loss": 0.9217,
      "step": 832
    },
    {
      "epoch": 0.33654865915862836,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0002802912621359223,
      "loss": 1.0137,
      "step": 833
    },
    {
      "epoch": 0.3369526791576183,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002802669902912621,
      "loss": 0.8218,
      "step": 834
    },
    {
      "epoch": 0.33735669915660826,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0002802427184466019,
      "loss": 0.8556,
      "step": 835
    },
    {
      "epoch": 0.3377607191555982,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00028021844660194173,
      "loss": 0.9511,
      "step": 836
    },
    {
      "epoch": 0.33816473915458817,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00028019417475728156,
      "loss": 0.9537,
      "step": 837
    },
    {
      "epoch": 0.3385687591535781,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00028016990291262133,
      "loss": 1.0392,
      "step": 838
    },
    {
      "epoch": 0.33897277915256807,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00028014563106796116,
      "loss": 0.9596,
      "step": 839
    },
    {
      "epoch": 0.339376799151558,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00028012135922330094,
      "loss": 1.0211,
      "step": 840
    },
    {
      "epoch": 0.339780819150548,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002800970873786407,
      "loss": 0.803,
      "step": 841
    },
    {
      "epoch": 0.3401848391495379,
      "grad_norm": 0.640625,
      "learning_rate": 0.00028007281553398054,
      "loss": 0.9732,
      "step": 842
    },
    {
      "epoch": 0.3405888591485279,
      "grad_norm": 0.65625,
      "learning_rate": 0.00028004854368932037,
      "loss": 1.0467,
      "step": 843
    },
    {
      "epoch": 0.3409928791475178,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0002800242718446602,
      "loss": 0.9668,
      "step": 844
    },
    {
      "epoch": 0.3413968991465077,
      "grad_norm": 0.609375,
      "learning_rate": 0.00028,
      "loss": 0.8563,
      "step": 845
    },
    {
      "epoch": 0.3418009191454977,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00027997572815533975,
      "loss": 0.9716,
      "step": 846
    },
    {
      "epoch": 0.34220493914448763,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002799514563106796,
      "loss": 0.9822,
      "step": 847
    },
    {
      "epoch": 0.3426089591434776,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002799271844660194,
      "loss": 0.8337,
      "step": 848
    },
    {
      "epoch": 0.34301297914246753,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002799029126213592,
      "loss": 0.8042,
      "step": 849
    },
    {
      "epoch": 0.3434169991414575,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000279878640776699,
      "loss": 0.9065,
      "step": 850
    },
    {
      "epoch": 0.34382101914044744,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002798543689320388,
      "loss": 0.9616,
      "step": 851
    },
    {
      "epoch": 0.3442250391394374,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002798300970873786,
      "loss": 0.8142,
      "step": 852
    },
    {
      "epoch": 0.34462905913842734,
      "grad_norm": 0.640625,
      "learning_rate": 0.00027980582524271845,
      "loss": 0.8965,
      "step": 853
    },
    {
      "epoch": 0.3450330791374173,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002797815533980582,
      "loss": 0.8453,
      "step": 854
    },
    {
      "epoch": 0.34543709913640724,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027975728155339805,
      "loss": 0.8563,
      "step": 855
    },
    {
      "epoch": 0.3458411191353972,
      "grad_norm": 0.6875,
      "learning_rate": 0.00027973300970873783,
      "loss": 0.9214,
      "step": 856
    },
    {
      "epoch": 0.34624513913438715,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00027970873786407766,
      "loss": 0.8165,
      "step": 857
    },
    {
      "epoch": 0.3466491591333771,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002796844660194175,
      "loss": 1.0236,
      "step": 858
    },
    {
      "epoch": 0.34705317913236705,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00027966019417475726,
      "loss": 0.8289,
      "step": 859
    },
    {
      "epoch": 0.347457199131357,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00027963592233009704,
      "loss": 0.8696,
      "step": 860
    },
    {
      "epoch": 0.34786121913034695,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027961165048543687,
      "loss": 0.9748,
      "step": 861
    },
    {
      "epoch": 0.3482652391293369,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002795873786407767,
      "loss": 0.9792,
      "step": 862
    },
    {
      "epoch": 0.34866925912832686,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00027956310679611647,
      "loss": 0.9177,
      "step": 863
    },
    {
      "epoch": 0.3490732791273168,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002795388349514563,
      "loss": 1.0146,
      "step": 864
    },
    {
      "epoch": 0.34947729912630676,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002795145631067961,
      "loss": 0.9266,
      "step": 865
    },
    {
      "epoch": 0.3498813191252967,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002794902912621359,
      "loss": 0.8002,
      "step": 866
    },
    {
      "epoch": 0.35028533912428667,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002794660194174757,
      "loss": 0.9918,
      "step": 867
    },
    {
      "epoch": 0.3506893591232766,
      "grad_norm": 0.70703125,
      "learning_rate": 0.0002794417475728155,
      "loss": 1.0639,
      "step": 868
    },
    {
      "epoch": 0.35109337912226657,
      "grad_norm": 0.671875,
      "learning_rate": 0.00027941747572815534,
      "loss": 0.9588,
      "step": 869
    },
    {
      "epoch": 0.3514973991212565,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002793932038834951,
      "loss": 0.7633,
      "step": 870
    },
    {
      "epoch": 0.3519014191202465,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002793689320388349,
      "loss": 0.896,
      "step": 871
    },
    {
      "epoch": 0.3523054391192364,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002793446601941747,
      "loss": 0.8646,
      "step": 872
    },
    {
      "epoch": 0.3527094591182264,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00027932038834951455,
      "loss": 0.8272,
      "step": 873
    },
    {
      "epoch": 0.3531134791172163,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002792961165048543,
      "loss": 0.9748,
      "step": 874
    },
    {
      "epoch": 0.3535174991162063,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00027927184466019415,
      "loss": 0.8959,
      "step": 875
    },
    {
      "epoch": 0.3539215191151962,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00027924757281553393,
      "loss": 0.8645,
      "step": 876
    },
    {
      "epoch": 0.3543255391141861,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00027922330097087376,
      "loss": 0.8721,
      "step": 877
    },
    {
      "epoch": 0.3547295591131761,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002791990291262136,
      "loss": 0.891,
      "step": 878
    },
    {
      "epoch": 0.35513357911216603,
      "grad_norm": 0.75,
      "learning_rate": 0.00027917475728155336,
      "loss": 0.9983,
      "step": 879
    },
    {
      "epoch": 0.355537599111156,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002791504854368932,
      "loss": 0.9142,
      "step": 880
    },
    {
      "epoch": 0.35594161911014593,
      "grad_norm": 0.546875,
      "learning_rate": 0.00027912621359223297,
      "loss": 0.8134,
      "step": 881
    },
    {
      "epoch": 0.3563456391091359,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002791019417475728,
      "loss": 0.826,
      "step": 882
    },
    {
      "epoch": 0.35674965910812584,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002790776699029126,
      "loss": 0.8607,
      "step": 883
    },
    {
      "epoch": 0.3571536791071158,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002790533980582524,
      "loss": 0.8213,
      "step": 884
    },
    {
      "epoch": 0.35755769910610574,
      "grad_norm": 0.5625,
      "learning_rate": 0.00027902912621359223,
      "loss": 0.8384,
      "step": 885
    },
    {
      "epoch": 0.3579617191050957,
      "grad_norm": 0.609375,
      "learning_rate": 0.000279004854368932,
      "loss": 1.02,
      "step": 886
    },
    {
      "epoch": 0.35836573910408565,
      "grad_norm": 0.53125,
      "learning_rate": 0.00027898058252427184,
      "loss": 0.8105,
      "step": 887
    },
    {
      "epoch": 0.3587697591030756,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002789563106796116,
      "loss": 0.8466,
      "step": 888
    },
    {
      "epoch": 0.35917377910206555,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00027893203883495144,
      "loss": 0.8651,
      "step": 889
    },
    {
      "epoch": 0.35957779910105553,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002789077669902912,
      "loss": 0.7706,
      "step": 890
    },
    {
      "epoch": 0.35998181910004545,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00027888349514563105,
      "loss": 0.8377,
      "step": 891
    },
    {
      "epoch": 0.3603858390990354,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002788592233009708,
      "loss": 0.8844,
      "step": 892
    },
    {
      "epoch": 0.36078985909802536,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027883495145631065,
      "loss": 0.9498,
      "step": 893
    },
    {
      "epoch": 0.3611938790970153,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002788106796116505,
      "loss": 0.9199,
      "step": 894
    },
    {
      "epoch": 0.36159789909600526,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00027878640776699026,
      "loss": 0.9291,
      "step": 895
    },
    {
      "epoch": 0.3620019190949952,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002787621359223301,
      "loss": 0.8296,
      "step": 896
    },
    {
      "epoch": 0.36240593909398516,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00027873786407766986,
      "loss": 0.8558,
      "step": 897
    },
    {
      "epoch": 0.3628099590929751,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002787135922330097,
      "loss": 0.9345,
      "step": 898
    },
    {
      "epoch": 0.36321397909196507,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002786893203883495,
      "loss": 0.9616,
      "step": 899
    },
    {
      "epoch": 0.363617999090955,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002786650485436893,
      "loss": 0.7173,
      "step": 900
    },
    {
      "epoch": 0.36402201908994497,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00027864077669902907,
      "loss": 0.9432,
      "step": 901
    },
    {
      "epoch": 0.3644260390889349,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002786165048543689,
      "loss": 0.8143,
      "step": 902
    },
    {
      "epoch": 0.3648300590879249,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00027859223300970873,
      "loss": 0.8476,
      "step": 903
    },
    {
      "epoch": 0.3652340790869148,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002785679611650485,
      "loss": 0.9911,
      "step": 904
    },
    {
      "epoch": 0.3656380990859048,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00027854368932038833,
      "loss": 0.7352,
      "step": 905
    },
    {
      "epoch": 0.3660421190848947,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002785194174757281,
      "loss": 0.8646,
      "step": 906
    },
    {
      "epoch": 0.3664461390838846,
      "grad_norm": 0.578125,
      "learning_rate": 0.00027849514563106794,
      "loss": 0.891,
      "step": 907
    },
    {
      "epoch": 0.3668501590828746,
      "grad_norm": 0.578125,
      "learning_rate": 0.00027847087378640777,
      "loss": 0.8848,
      "step": 908
    },
    {
      "epoch": 0.36725417908186453,
      "grad_norm": 0.640625,
      "learning_rate": 0.00027844660194174754,
      "loss": 1.0004,
      "step": 909
    },
    {
      "epoch": 0.3676581990808545,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00027842233009708737,
      "loss": 1.0562,
      "step": 910
    },
    {
      "epoch": 0.36806221907984443,
      "grad_norm": 0.671875,
      "learning_rate": 0.00027839805825242715,
      "loss": 0.9096,
      "step": 911
    },
    {
      "epoch": 0.3684662390788344,
      "grad_norm": 0.55078125,
      "learning_rate": 0.000278373786407767,
      "loss": 0.8454,
      "step": 912
    },
    {
      "epoch": 0.36887025907782434,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002783495145631068,
      "loss": 0.8723,
      "step": 913
    },
    {
      "epoch": 0.3692742790768143,
      "grad_norm": 0.671875,
      "learning_rate": 0.0002783252427184466,
      "loss": 0.9513,
      "step": 914
    },
    {
      "epoch": 0.36967829907580424,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002783009708737864,
      "loss": 0.9708,
      "step": 915
    },
    {
      "epoch": 0.3700823190747942,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0002782766990291262,
      "loss": 0.9693,
      "step": 916
    },
    {
      "epoch": 0.37048633907378414,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00027825242718446596,
      "loss": 0.8826,
      "step": 917
    },
    {
      "epoch": 0.3708903590727741,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002782281553398058,
      "loss": 0.9152,
      "step": 918
    },
    {
      "epoch": 0.37129437907176405,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002782038834951456,
      "loss": 0.9224,
      "step": 919
    },
    {
      "epoch": 0.371698399070754,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002781796116504854,
      "loss": 0.8963,
      "step": 920
    },
    {
      "epoch": 0.37210241906974395,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002781553398058252,
      "loss": 0.8334,
      "step": 921
    },
    {
      "epoch": 0.37250643906873393,
      "grad_norm": 0.53515625,
      "learning_rate": 0.000278131067961165,
      "loss": 0.7856,
      "step": 922
    },
    {
      "epoch": 0.37291045906772385,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027810679611650483,
      "loss": 0.9035,
      "step": 923
    },
    {
      "epoch": 0.3733144790667138,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00027808252427184466,
      "loss": 0.9386,
      "step": 924
    },
    {
      "epoch": 0.37371849906570376,
      "grad_norm": 0.609375,
      "learning_rate": 0.00027805825242718443,
      "loss": 0.8479,
      "step": 925
    },
    {
      "epoch": 0.3741225190646937,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00027803398058252426,
      "loss": 0.8504,
      "step": 926
    },
    {
      "epoch": 0.37452653906368366,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00027800970873786404,
      "loss": 0.7408,
      "step": 927
    },
    {
      "epoch": 0.3749305590626736,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00027798543689320387,
      "loss": 0.871,
      "step": 928
    },
    {
      "epoch": 0.37533457906166356,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002779611650485437,
      "loss": 0.9568,
      "step": 929
    },
    {
      "epoch": 0.3757385990606535,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00027793689320388347,
      "loss": 0.8517,
      "step": 930
    },
    {
      "epoch": 0.37614261905964347,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027791262135922325,
      "loss": 0.8995,
      "step": 931
    },
    {
      "epoch": 0.3765466390586334,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002778883495145631,
      "loss": 0.8633,
      "step": 932
    },
    {
      "epoch": 0.37695065905762337,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002778640776699029,
      "loss": 0.7833,
      "step": 933
    },
    {
      "epoch": 0.3773546790566133,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002778398058252427,
      "loss": 0.86,
      "step": 934
    },
    {
      "epoch": 0.3777586990556033,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002778155339805825,
      "loss": 0.9247,
      "step": 935
    },
    {
      "epoch": 0.3781627190545932,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002777912621359223,
      "loss": 0.8955,
      "step": 936
    },
    {
      "epoch": 0.3785667390535832,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002777669902912621,
      "loss": 0.9596,
      "step": 937
    },
    {
      "epoch": 0.3789707590525731,
      "grad_norm": 0.671875,
      "learning_rate": 0.00027774271844660195,
      "loss": 0.9683,
      "step": 938
    },
    {
      "epoch": 0.379374779051563,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002777184466019417,
      "loss": 0.8889,
      "step": 939
    },
    {
      "epoch": 0.379778799050553,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00027769417475728155,
      "loss": 0.8788,
      "step": 940
    },
    {
      "epoch": 0.38018281904954293,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002776699029126213,
      "loss": 0.8881,
      "step": 941
    },
    {
      "epoch": 0.3805868390485329,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002776456310679611,
      "loss": 0.9409,
      "step": 942
    },
    {
      "epoch": 0.38099085904752283,
      "grad_norm": 0.515625,
      "learning_rate": 0.00027762135922330093,
      "loss": 0.8313,
      "step": 943
    },
    {
      "epoch": 0.3813948790465128,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00027759708737864076,
      "loss": 0.8661,
      "step": 944
    },
    {
      "epoch": 0.38179889904550274,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002775728155339806,
      "loss": 0.9431,
      "step": 945
    },
    {
      "epoch": 0.3822029190444927,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00027754854368932036,
      "loss": 0.8562,
      "step": 946
    },
    {
      "epoch": 0.38260693904348264,
      "grad_norm": 0.609375,
      "learning_rate": 0.00027752427184466014,
      "loss": 0.8767,
      "step": 947
    },
    {
      "epoch": 0.3830109590424726,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00027749999999999997,
      "loss": 0.9938,
      "step": 948
    },
    {
      "epoch": 0.38341497904146254,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002774757281553398,
      "loss": 0.895,
      "step": 949
    },
    {
      "epoch": 0.3838189990404525,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002774514563106796,
      "loss": 0.8302,
      "step": 950
    },
    {
      "epoch": 0.38422301903944245,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002774271844660194,
      "loss": 0.9076,
      "step": 951
    },
    {
      "epoch": 0.38462703903843243,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002774029126213592,
      "loss": 0.8623,
      "step": 952
    },
    {
      "epoch": 0.38503105903742235,
      "grad_norm": 0.5234375,
      "learning_rate": 0.000277378640776699,
      "loss": 0.874,
      "step": 953
    },
    {
      "epoch": 0.3854350790364123,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00027735436893203884,
      "loss": 0.9533,
      "step": 954
    },
    {
      "epoch": 0.38583909903540226,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002773300970873786,
      "loss": 1.0447,
      "step": 955
    },
    {
      "epoch": 0.3862431190343922,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00027730582524271844,
      "loss": 0.7848,
      "step": 956
    },
    {
      "epoch": 0.38664713903338216,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002772815533980582,
      "loss": 0.8539,
      "step": 957
    },
    {
      "epoch": 0.3870511590323721,
      "grad_norm": 0.71875,
      "learning_rate": 0.00027725728155339805,
      "loss": 0.9533,
      "step": 958
    },
    {
      "epoch": 0.38745517903136206,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002772330097087379,
      "loss": 0.8743,
      "step": 959
    },
    {
      "epoch": 0.387859199030352,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027720873786407765,
      "loss": 0.8219,
      "step": 960
    },
    {
      "epoch": 0.38826321902934197,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002771844660194174,
      "loss": 0.9349,
      "step": 961
    },
    {
      "epoch": 0.3886672390283319,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00027716019417475726,
      "loss": 0.888,
      "step": 962
    },
    {
      "epoch": 0.38907125902732187,
      "grad_norm": 2.84375,
      "learning_rate": 0.0002771359223300971,
      "loss": 0.9392,
      "step": 963
    },
    {
      "epoch": 0.3894752790263118,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00027711165048543686,
      "loss": 0.9082,
      "step": 964
    },
    {
      "epoch": 0.3898792990253018,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002770873786407767,
      "loss": 0.884,
      "step": 965
    },
    {
      "epoch": 0.3902833190242917,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00027706310679611647,
      "loss": 0.9095,
      "step": 966
    },
    {
      "epoch": 0.3906873390232817,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002770388349514563,
      "loss": 0.9312,
      "step": 967
    },
    {
      "epoch": 0.3910913590222716,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00027701456310679607,
      "loss": 0.9183,
      "step": 968
    },
    {
      "epoch": 0.3914953790212615,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002769902912621359,
      "loss": 0.9443,
      "step": 969
    },
    {
      "epoch": 0.3918993990202515,
      "grad_norm": 0.5625,
      "learning_rate": 0.00027696601941747573,
      "loss": 0.896,
      "step": 970
    },
    {
      "epoch": 0.39230341901924143,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002769417475728155,
      "loss": 0.8419,
      "step": 971
    },
    {
      "epoch": 0.3927074390182314,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0002769174757281553,
      "loss": 0.9442,
      "step": 972
    },
    {
      "epoch": 0.39311145901722133,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002768932038834951,
      "loss": 0.9117,
      "step": 973
    },
    {
      "epoch": 0.3935154790162113,
      "grad_norm": 0.578125,
      "learning_rate": 0.00027686893203883494,
      "loss": 0.8801,
      "step": 974
    },
    {
      "epoch": 0.39391949901520124,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002768446601941747,
      "loss": 0.913,
      "step": 975
    },
    {
      "epoch": 0.3943235190141912,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00027682038834951454,
      "loss": 0.9184,
      "step": 976
    },
    {
      "epoch": 0.39472753901318114,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002767961165048543,
      "loss": 0.8952,
      "step": 977
    },
    {
      "epoch": 0.3951315590121711,
      "grad_norm": 0.609375,
      "learning_rate": 0.00027677184466019415,
      "loss": 0.824,
      "step": 978
    },
    {
      "epoch": 0.39553557901116104,
      "grad_norm": 0.5625,
      "learning_rate": 0.000276747572815534,
      "loss": 0.8904,
      "step": 979
    },
    {
      "epoch": 0.395939599010151,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00027672330097087375,
      "loss": 0.8198,
      "step": 980
    },
    {
      "epoch": 0.39634361900914095,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0002766990291262136,
      "loss": 0.9305,
      "step": 981
    },
    {
      "epoch": 0.3967476390081309,
      "grad_norm": 0.578125,
      "learning_rate": 0.00027667475728155336,
      "loss": 0.877,
      "step": 982
    },
    {
      "epoch": 0.39715165900712085,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002766504854368932,
      "loss": 0.881,
      "step": 983
    },
    {
      "epoch": 0.39755567900611083,
      "grad_norm": 0.65625,
      "learning_rate": 0.000276626213592233,
      "loss": 0.9868,
      "step": 984
    },
    {
      "epoch": 0.39795969900510075,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002766019417475728,
      "loss": 0.9481,
      "step": 985
    },
    {
      "epoch": 0.3983637190040907,
      "grad_norm": 2.5625,
      "learning_rate": 0.0002765776699029126,
      "loss": 0.8312,
      "step": 986
    },
    {
      "epoch": 0.39876773900308066,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002765533980582524,
      "loss": 0.8199,
      "step": 987
    },
    {
      "epoch": 0.3991717590020706,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0002765291262135922,
      "loss": 0.9141,
      "step": 988
    },
    {
      "epoch": 0.39957577900106056,
      "grad_norm": 0.55078125,
      "learning_rate": 0.000276504854368932,
      "loss": 0.9191,
      "step": 989
    },
    {
      "epoch": 0.3999797990000505,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00027648058252427183,
      "loss": 0.7712,
      "step": 990
    },
    {
      "epoch": 0.40038381899904046,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002764563106796116,
      "loss": 0.9265,
      "step": 991
    },
    {
      "epoch": 0.4007878389980304,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00027643203883495143,
      "loss": 0.8154,
      "step": 992
    },
    {
      "epoch": 0.40119185899702037,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002764077669902912,
      "loss": 0.9648,
      "step": 993
    },
    {
      "epoch": 0.4015958789960103,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00027638349514563104,
      "loss": 0.8854,
      "step": 994
    },
    {
      "epoch": 0.40199989899500027,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00027635922330097087,
      "loss": 0.9464,
      "step": 995
    },
    {
      "epoch": 0.4024039189939902,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00027633495145631064,
      "loss": 0.9086,
      "step": 996
    },
    {
      "epoch": 0.4028079389929802,
      "grad_norm": 0.8125,
      "learning_rate": 0.0002763106796116505,
      "loss": 1.0486,
      "step": 997
    },
    {
      "epoch": 0.4032119589919701,
      "grad_norm": 0.609375,
      "learning_rate": 0.00027628640776699025,
      "loss": 0.9475,
      "step": 998
    },
    {
      "epoch": 0.4036159789909601,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002762621359223301,
      "loss": 0.985,
      "step": 999
    },
    {
      "epoch": 0.40401999898995,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0002762378640776699,
      "loss": 0.9732,
      "step": 1000
    },
    {
      "epoch": 0.4044240189889399,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002762135922330097,
      "loss": 0.9496,
      "step": 1001
    },
    {
      "epoch": 0.4048280389879299,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00027618932038834946,
      "loss": 0.798,
      "step": 1002
    },
    {
      "epoch": 0.40523205898691983,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002761650485436893,
      "loss": 0.9126,
      "step": 1003
    },
    {
      "epoch": 0.4056360789859098,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002761407766990291,
      "loss": 0.8748,
      "step": 1004
    },
    {
      "epoch": 0.40604009898489973,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002761165048543689,
      "loss": 0.8633,
      "step": 1005
    },
    {
      "epoch": 0.4064441189838897,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002760922330097087,
      "loss": 0.928,
      "step": 1006
    },
    {
      "epoch": 0.40684813898287964,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002760679611650485,
      "loss": 0.8375,
      "step": 1007
    },
    {
      "epoch": 0.4072521589818696,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002760436893203883,
      "loss": 0.9303,
      "step": 1008
    },
    {
      "epoch": 0.40765617898085954,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00027601941747572816,
      "loss": 0.8973,
      "step": 1009
    },
    {
      "epoch": 0.4080601989798495,
      "grad_norm": 0.609375,
      "learning_rate": 0.00027599514563106793,
      "loss": 0.8821,
      "step": 1010
    },
    {
      "epoch": 0.40846421897883944,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00027597087378640776,
      "loss": 0.8353,
      "step": 1011
    },
    {
      "epoch": 0.4088682389778294,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00027594660194174754,
      "loss": 1.0146,
      "step": 1012
    },
    {
      "epoch": 0.40927225897681935,
      "grad_norm": 1.453125,
      "learning_rate": 0.00027592233009708736,
      "loss": 0.8854,
      "step": 1013
    },
    {
      "epoch": 0.4096762789758093,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002758980582524272,
      "loss": 0.8582,
      "step": 1014
    },
    {
      "epoch": 0.41008029897479925,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00027587378640776697,
      "loss": 0.8418,
      "step": 1015
    },
    {
      "epoch": 0.4104843189737892,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002758495145631068,
      "loss": 0.8894,
      "step": 1016
    },
    {
      "epoch": 0.41088833897277915,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002758252427184466,
      "loss": 0.948,
      "step": 1017
    },
    {
      "epoch": 0.4112923589717691,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00027580097087378635,
      "loss": 0.8727,
      "step": 1018
    },
    {
      "epoch": 0.41169637897075906,
      "grad_norm": 0.671875,
      "learning_rate": 0.0002757766990291262,
      "loss": 1.0001,
      "step": 1019
    },
    {
      "epoch": 0.412100398969749,
      "grad_norm": 0.5859375,
      "learning_rate": 0.000275752427184466,
      "loss": 0.8405,
      "step": 1020
    },
    {
      "epoch": 0.41250441896873896,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002757281553398058,
      "loss": 0.9275,
      "step": 1021
    },
    {
      "epoch": 0.4129084389677289,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002757038834951456,
      "loss": 0.9575,
      "step": 1022
    },
    {
      "epoch": 0.41331245896671887,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002756796116504854,
      "loss": 0.8439,
      "step": 1023
    },
    {
      "epoch": 0.4137164789657088,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002756553398058252,
      "loss": 0.8572,
      "step": 1024
    },
    {
      "epoch": 0.41412049896469877,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00027563106796116505,
      "loss": 0.842,
      "step": 1025
    },
    {
      "epoch": 0.4145245189636887,
      "grad_norm": 0.625,
      "learning_rate": 0.0002756067961165048,
      "loss": 0.854,
      "step": 1026
    },
    {
      "epoch": 0.4149285389626787,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00027558252427184465,
      "loss": 0.921,
      "step": 1027
    },
    {
      "epoch": 0.4153325589616686,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00027555825242718443,
      "loss": 0.7984,
      "step": 1028
    },
    {
      "epoch": 0.4157365789606586,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00027553398058252426,
      "loss": 0.912,
      "step": 1029
    },
    {
      "epoch": 0.4161405989596485,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002755097087378641,
      "loss": 0.8318,
      "step": 1030
    },
    {
      "epoch": 0.4165446189586385,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00027548543689320386,
      "loss": 0.8686,
      "step": 1031
    },
    {
      "epoch": 0.4169486389576284,
      "grad_norm": 0.625,
      "learning_rate": 0.00027546116504854364,
      "loss": 0.9132,
      "step": 1032
    },
    {
      "epoch": 0.4173526589566183,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027543689320388347,
      "loss": 0.8349,
      "step": 1033
    },
    {
      "epoch": 0.4177566789556083,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002754126213592233,
      "loss": 0.8891,
      "step": 1034
    },
    {
      "epoch": 0.41816069895459823,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00027538834951456307,
      "loss": 0.98,
      "step": 1035
    },
    {
      "epoch": 0.4185647189535882,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002753640776699029,
      "loss": 0.8228,
      "step": 1036
    },
    {
      "epoch": 0.41896873895257813,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002753398058252427,
      "loss": 0.9978,
      "step": 1037
    },
    {
      "epoch": 0.4193727589515681,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002753155339805825,
      "loss": 0.82,
      "step": 1038
    },
    {
      "epoch": 0.41977677895055804,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00027529126213592233,
      "loss": 0.9223,
      "step": 1039
    },
    {
      "epoch": 0.420180798949548,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002752669902912621,
      "loss": 0.9154,
      "step": 1040
    },
    {
      "epoch": 0.42058481894853794,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027524271844660194,
      "loss": 0.988,
      "step": 1041
    },
    {
      "epoch": 0.4209888389475279,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002752184466019417,
      "loss": 0.8191,
      "step": 1042
    },
    {
      "epoch": 0.42139285894651785,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002751941747572815,
      "loss": 1.0068,
      "step": 1043
    },
    {
      "epoch": 0.4217968789455078,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002751699029126213,
      "loss": 0.907,
      "step": 1044
    },
    {
      "epoch": 0.42220089894449775,
      "grad_norm": 0.515625,
      "learning_rate": 0.00027514563106796115,
      "loss": 0.7958,
      "step": 1045
    },
    {
      "epoch": 0.42260491894348773,
      "grad_norm": 0.625,
      "learning_rate": 0.000275121359223301,
      "loss": 0.8603,
      "step": 1046
    },
    {
      "epoch": 0.42300893894247765,
      "grad_norm": 0.53125,
      "learning_rate": 0.00027509708737864075,
      "loss": 0.821,
      "step": 1047
    },
    {
      "epoch": 0.4234129589414676,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00027507281553398053,
      "loss": 0.8767,
      "step": 1048
    },
    {
      "epoch": 0.42381697894045756,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00027504854368932036,
      "loss": 0.9093,
      "step": 1049
    },
    {
      "epoch": 0.4242209989394475,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002750242718446602,
      "loss": 0.7834,
      "step": 1050
    },
    {
      "epoch": 0.42462501893843746,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00027499999999999996,
      "loss": 0.9573,
      "step": 1051
    },
    {
      "epoch": 0.4250290389374274,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0002749757281553398,
      "loss": 0.9883,
      "step": 1052
    },
    {
      "epoch": 0.42543305893641736,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00027495145631067957,
      "loss": 0.8401,
      "step": 1053
    },
    {
      "epoch": 0.4258370789354073,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0002749271844660194,
      "loss": 0.9947,
      "step": 1054
    },
    {
      "epoch": 0.42624109893439727,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002749029126213592,
      "loss": 0.8611,
      "step": 1055
    },
    {
      "epoch": 0.4266451189333872,
      "grad_norm": 0.984375,
      "learning_rate": 0.000274878640776699,
      "loss": 0.9094,
      "step": 1056
    },
    {
      "epoch": 0.42704913893237717,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027485436893203883,
      "loss": 0.9064,
      "step": 1057
    },
    {
      "epoch": 0.4274531589313671,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002748300970873786,
      "loss": 0.924,
      "step": 1058
    },
    {
      "epoch": 0.4278571789303571,
      "grad_norm": 0.53125,
      "learning_rate": 0.00027480582524271844,
      "loss": 0.866,
      "step": 1059
    },
    {
      "epoch": 0.428261198929347,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027478155339805826,
      "loss": 0.9385,
      "step": 1060
    },
    {
      "epoch": 0.428665218928337,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00027475728155339804,
      "loss": 0.7783,
      "step": 1061
    },
    {
      "epoch": 0.4290692389273269,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002747330097087378,
      "loss": 1.1032,
      "step": 1062
    },
    {
      "epoch": 0.4294732589263168,
      "grad_norm": 0.5625,
      "learning_rate": 0.00027470873786407764,
      "loss": 0.847,
      "step": 1063
    },
    {
      "epoch": 0.4298772789253068,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002746844660194175,
      "loss": 0.8389,
      "step": 1064
    },
    {
      "epoch": 0.43028129892429673,
      "grad_norm": 0.625,
      "learning_rate": 0.00027466019417475725,
      "loss": 0.9199,
      "step": 1065
    },
    {
      "epoch": 0.4306853189232867,
      "grad_norm": 0.703125,
      "learning_rate": 0.0002746359223300971,
      "loss": 0.9854,
      "step": 1066
    },
    {
      "epoch": 0.43108933892227663,
      "grad_norm": 0.78125,
      "learning_rate": 0.00027461165048543685,
      "loss": 1.0216,
      "step": 1067
    },
    {
      "epoch": 0.4314933589212666,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002745873786407767,
      "loss": 0.8628,
      "step": 1068
    },
    {
      "epoch": 0.43189737892025654,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00027456310679611646,
      "loss": 0.7613,
      "step": 1069
    },
    {
      "epoch": 0.4323013989192465,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002745388349514563,
      "loss": 0.9457,
      "step": 1070
    },
    {
      "epoch": 0.43270541891823644,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002745145631067961,
      "loss": 0.8058,
      "step": 1071
    },
    {
      "epoch": 0.4331094389172264,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002744902912621359,
      "loss": 0.8303,
      "step": 1072
    },
    {
      "epoch": 0.43351345891621634,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00027446601941747567,
      "loss": 0.7541,
      "step": 1073
    },
    {
      "epoch": 0.4339174789152063,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002744417475728155,
      "loss": 0.8887,
      "step": 1074
    },
    {
      "epoch": 0.43432149891419625,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002744174757281553,
      "loss": 0.9178,
      "step": 1075
    },
    {
      "epoch": 0.4347255189131862,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002743932038834951,
      "loss": 0.9269,
      "step": 1076
    },
    {
      "epoch": 0.43512953891217615,
      "grad_norm": 0.53125,
      "learning_rate": 0.00027436893203883493,
      "loss": 0.8439,
      "step": 1077
    },
    {
      "epoch": 0.43553355891116613,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002743446601941747,
      "loss": 0.92,
      "step": 1078
    },
    {
      "epoch": 0.43593757891015605,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00027432038834951454,
      "loss": 0.9346,
      "step": 1079
    },
    {
      "epoch": 0.436341598909146,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00027429611650485437,
      "loss": 0.8247,
      "step": 1080
    },
    {
      "epoch": 0.43674561890813596,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00027427184466019414,
      "loss": 0.7588,
      "step": 1081
    },
    {
      "epoch": 0.4371496389071259,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00027424757281553397,
      "loss": 0.9359,
      "step": 1082
    },
    {
      "epoch": 0.43755365890611586,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00027422330097087375,
      "loss": 0.9474,
      "step": 1083
    },
    {
      "epoch": 0.4379576789051058,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002741990291262136,
      "loss": 0.8526,
      "step": 1084
    },
    {
      "epoch": 0.43836169890409576,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002741747572815534,
      "loss": 0.9287,
      "step": 1085
    },
    {
      "epoch": 0.4387657189030857,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002741504854368932,
      "loss": 0.8832,
      "step": 1086
    },
    {
      "epoch": 0.43916973890207567,
      "grad_norm": 0.66796875,
      "learning_rate": 0.000274126213592233,
      "loss": 0.9876,
      "step": 1087
    },
    {
      "epoch": 0.4395737589010656,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002741019417475728,
      "loss": 0.8628,
      "step": 1088
    },
    {
      "epoch": 0.43997777890005557,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002740776699029126,
      "loss": 0.8763,
      "step": 1089
    },
    {
      "epoch": 0.4403817988990455,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002740533980582524,
      "loss": 0.8421,
      "step": 1090
    },
    {
      "epoch": 0.4407858188980355,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002740291262135922,
      "loss": 0.8245,
      "step": 1091
    },
    {
      "epoch": 0.4411898388970254,
      "grad_norm": 0.6640625,
      "learning_rate": 0.000274004854368932,
      "loss": 0.9276,
      "step": 1092
    },
    {
      "epoch": 0.4415938588960154,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002739805825242718,
      "loss": 0.8337,
      "step": 1093
    },
    {
      "epoch": 0.4419978788950053,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0002739563106796116,
      "loss": 1.0179,
      "step": 1094
    },
    {
      "epoch": 0.4424018988939952,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00027393203883495143,
      "loss": 0.8817,
      "step": 1095
    },
    {
      "epoch": 0.4428059188929852,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00027390776699029126,
      "loss": 0.9082,
      "step": 1096
    },
    {
      "epoch": 0.44320993889197513,
      "grad_norm": 0.546875,
      "learning_rate": 0.00027388349514563103,
      "loss": 0.7794,
      "step": 1097
    },
    {
      "epoch": 0.4436139588909651,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00027385922330097086,
      "loss": 0.7876,
      "step": 1098
    },
    {
      "epoch": 0.44401797888995503,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00027383495145631064,
      "loss": 0.8368,
      "step": 1099
    },
    {
      "epoch": 0.444421998888945,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00027381067961165047,
      "loss": 0.9497,
      "step": 1100
    },
    {
      "epoch": 0.44482601888793494,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002737864077669903,
      "loss": 0.9198,
      "step": 1101
    },
    {
      "epoch": 0.4452300388869249,
      "grad_norm": 0.5625,
      "learning_rate": 0.00027376213592233007,
      "loss": 0.8391,
      "step": 1102
    },
    {
      "epoch": 0.44563405888591484,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00027373786407766985,
      "loss": 0.8131,
      "step": 1103
    },
    {
      "epoch": 0.4460380788849048,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002737135922330097,
      "loss": 0.8853,
      "step": 1104
    },
    {
      "epoch": 0.44644209888389474,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0002736893203883495,
      "loss": 0.8473,
      "step": 1105
    },
    {
      "epoch": 0.4468461188828847,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002736650485436893,
      "loss": 0.9862,
      "step": 1106
    },
    {
      "epoch": 0.44725013888187465,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002736407766990291,
      "loss": 0.802,
      "step": 1107
    },
    {
      "epoch": 0.4476541588808646,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002736165048543689,
      "loss": 0.9025,
      "step": 1108
    },
    {
      "epoch": 0.44805817887985455,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002735922330097087,
      "loss": 0.9466,
      "step": 1109
    },
    {
      "epoch": 0.4484621988788445,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00027356796116504854,
      "loss": 0.7939,
      "step": 1110
    },
    {
      "epoch": 0.44886621887783446,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002735436893203883,
      "loss": 0.8642,
      "step": 1111
    },
    {
      "epoch": 0.4492702388768244,
      "grad_norm": 0.609375,
      "learning_rate": 0.00027351941747572815,
      "loss": 0.9476,
      "step": 1112
    },
    {
      "epoch": 0.44967425887581436,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002734951456310679,
      "loss": 0.9102,
      "step": 1113
    },
    {
      "epoch": 0.4500782788748043,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00027347087378640775,
      "loss": 0.8822,
      "step": 1114
    },
    {
      "epoch": 0.45048229887379426,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002734466019417476,
      "loss": 0.8893,
      "step": 1115
    },
    {
      "epoch": 0.4508863188727842,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00027342233009708736,
      "loss": 0.8921,
      "step": 1116
    },
    {
      "epoch": 0.45129033887177417,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002733980582524272,
      "loss": 0.8277,
      "step": 1117
    },
    {
      "epoch": 0.4516943588707641,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00027337378640776696,
      "loss": 0.8439,
      "step": 1118
    },
    {
      "epoch": 0.45209837886975407,
      "grad_norm": 0.6875,
      "learning_rate": 0.00027334951456310674,
      "loss": 1.0,
      "step": 1119
    },
    {
      "epoch": 0.452502398868744,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00027332524271844657,
      "loss": 0.946,
      "step": 1120
    },
    {
      "epoch": 0.452906418867734,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002733009708737864,
      "loss": 1.0198,
      "step": 1121
    },
    {
      "epoch": 0.4533104388667239,
      "grad_norm": 0.75,
      "learning_rate": 0.00027327669902912617,
      "loss": 0.9032,
      "step": 1122
    },
    {
      "epoch": 0.4537144588657139,
      "grad_norm": 0.66796875,
      "learning_rate": 0.000273252427184466,
      "loss": 0.9265,
      "step": 1123
    },
    {
      "epoch": 0.4541184788647038,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002732281553398058,
      "loss": 0.8998,
      "step": 1124
    },
    {
      "epoch": 0.4545224988636938,
      "grad_norm": 0.671875,
      "learning_rate": 0.0002732038834951456,
      "loss": 1.0331,
      "step": 1125
    },
    {
      "epoch": 0.4549265188626837,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027317961165048544,
      "loss": 0.8052,
      "step": 1126
    },
    {
      "epoch": 0.45533053886167363,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002731553398058252,
      "loss": 0.8508,
      "step": 1127
    },
    {
      "epoch": 0.4557345588606636,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00027313106796116504,
      "loss": 0.834,
      "step": 1128
    },
    {
      "epoch": 0.45613857885965353,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002731067961165048,
      "loss": 0.852,
      "step": 1129
    },
    {
      "epoch": 0.4565425988586435,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00027308252427184465,
      "loss": 0.9022,
      "step": 1130
    },
    {
      "epoch": 0.45694661885763344,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002730582524271845,
      "loss": 0.9075,
      "step": 1131
    },
    {
      "epoch": 0.4573506388566234,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00027303398058252425,
      "loss": 0.9308,
      "step": 1132
    },
    {
      "epoch": 0.45775465885561334,
      "grad_norm": 0.578125,
      "learning_rate": 0.000273009708737864,
      "loss": 0.9852,
      "step": 1133
    },
    {
      "epoch": 0.4581586788546033,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00027298543689320385,
      "loss": 0.8516,
      "step": 1134
    },
    {
      "epoch": 0.45856269885359324,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002729611650485437,
      "loss": 0.8424,
      "step": 1135
    },
    {
      "epoch": 0.4589667188525832,
      "grad_norm": 0.5,
      "learning_rate": 0.00027293689320388346,
      "loss": 0.8186,
      "step": 1136
    },
    {
      "epoch": 0.45937073885157315,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002729126213592233,
      "loss": 0.9224,
      "step": 1137
    },
    {
      "epoch": 0.4597747588505631,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00027288834951456306,
      "loss": 0.8946,
      "step": 1138
    },
    {
      "epoch": 0.46017877884955305,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002728640776699029,
      "loss": 0.9042,
      "step": 1139
    },
    {
      "epoch": 0.46058279884854303,
      "grad_norm": 0.484375,
      "learning_rate": 0.0002728398058252427,
      "loss": 0.8686,
      "step": 1140
    },
    {
      "epoch": 0.46098681884753295,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0002728155339805825,
      "loss": 1.0408,
      "step": 1141
    },
    {
      "epoch": 0.4613908388465229,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00027279126213592233,
      "loss": 0.93,
      "step": 1142
    },
    {
      "epoch": 0.46179485884551286,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002727669902912621,
      "loss": 0.8782,
      "step": 1143
    },
    {
      "epoch": 0.4621988788445028,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002727427184466019,
      "loss": 1.1381,
      "step": 1144
    },
    {
      "epoch": 0.46260289884349276,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002727184466019417,
      "loss": 0.7424,
      "step": 1145
    },
    {
      "epoch": 0.4630069188424827,
      "grad_norm": 0.65625,
      "learning_rate": 0.00027269417475728154,
      "loss": 0.9146,
      "step": 1146
    },
    {
      "epoch": 0.46341093884147266,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00027266990291262137,
      "loss": 0.8364,
      "step": 1147
    },
    {
      "epoch": 0.4638149588404626,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00027264563106796114,
      "loss": 0.9727,
      "step": 1148
    },
    {
      "epoch": 0.46421897883945257,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002726213592233009,
      "loss": 0.8953,
      "step": 1149
    },
    {
      "epoch": 0.4646229988384425,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00027259708737864075,
      "loss": 0.8648,
      "step": 1150
    },
    {
      "epoch": 0.46502701883743247,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002725728155339806,
      "loss": 0.9128,
      "step": 1151
    },
    {
      "epoch": 0.4654310388364224,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00027254854368932035,
      "loss": 0.8492,
      "step": 1152
    },
    {
      "epoch": 0.4658350588354124,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002725242718446602,
      "loss": 0.8966,
      "step": 1153
    },
    {
      "epoch": 0.4662390788344023,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00027249999999999996,
      "loss": 0.8473,
      "step": 1154
    },
    {
      "epoch": 0.4666430988333923,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002724757281553398,
      "loss": 0.8258,
      "step": 1155
    },
    {
      "epoch": 0.4670471188323822,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002724514563106796,
      "loss": 0.8523,
      "step": 1156
    },
    {
      "epoch": 0.4674511388313721,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0002724271844660194,
      "loss": 0.8554,
      "step": 1157
    },
    {
      "epoch": 0.4678551588303621,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002724029126213592,
      "loss": 0.8096,
      "step": 1158
    },
    {
      "epoch": 0.46825917882935203,
      "grad_norm": 0.6171875,
      "learning_rate": 0.000272378640776699,
      "loss": 0.8683,
      "step": 1159
    },
    {
      "epoch": 0.468663198828342,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002723543689320388,
      "loss": 0.8513,
      "step": 1160
    },
    {
      "epoch": 0.46906721882733193,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00027233009708737865,
      "loss": 0.7921,
      "step": 1161
    },
    {
      "epoch": 0.4694712388263219,
      "grad_norm": 1.9375,
      "learning_rate": 0.00027230582524271843,
      "loss": 1.013,
      "step": 1162
    },
    {
      "epoch": 0.46987525882531184,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002722815533980582,
      "loss": 0.8156,
      "step": 1163
    },
    {
      "epoch": 0.4702792788243018,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00027225728155339803,
      "loss": 0.9932,
      "step": 1164
    },
    {
      "epoch": 0.47068329882329174,
      "grad_norm": 0.640625,
      "learning_rate": 0.00027223300970873786,
      "loss": 0.9747,
      "step": 1165
    },
    {
      "epoch": 0.4710873188222817,
      "grad_norm": 0.65625,
      "learning_rate": 0.00027220873786407764,
      "loss": 0.9227,
      "step": 1166
    },
    {
      "epoch": 0.47149133882127164,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00027218446601941747,
      "loss": 0.8498,
      "step": 1167
    },
    {
      "epoch": 0.4718953588202616,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00027216019417475724,
      "loss": 0.8587,
      "step": 1168
    },
    {
      "epoch": 0.47229937881925155,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00027213592233009707,
      "loss": 0.8981,
      "step": 1169
    },
    {
      "epoch": 0.4727033988182415,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00027211165048543685,
      "loss": 0.8665,
      "step": 1170
    },
    {
      "epoch": 0.47310741881723145,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002720873786407767,
      "loss": 0.8903,
      "step": 1171
    },
    {
      "epoch": 0.47351143881622143,
      "grad_norm": 0.70703125,
      "learning_rate": 0.0002720631067961165,
      "loss": 0.914,
      "step": 1172
    },
    {
      "epoch": 0.47391545881521135,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002720388349514563,
      "loss": 0.8625,
      "step": 1173
    },
    {
      "epoch": 0.4743194788142013,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00027201456310679606,
      "loss": 0.8259,
      "step": 1174
    },
    {
      "epoch": 0.47472349881319126,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002719902912621359,
      "loss": 0.9212,
      "step": 1175
    },
    {
      "epoch": 0.4751275188121812,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002719660194174757,
      "loss": 0.8988,
      "step": 1176
    },
    {
      "epoch": 0.47553153881117116,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002719417475728155,
      "loss": 0.9419,
      "step": 1177
    },
    {
      "epoch": 0.4759355588101611,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002719174757281553,
      "loss": 0.9106,
      "step": 1178
    },
    {
      "epoch": 0.47633957880915107,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002718932038834951,
      "loss": 0.9255,
      "step": 1179
    },
    {
      "epoch": 0.476743598808141,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002718689320388349,
      "loss": 0.9378,
      "step": 1180
    },
    {
      "epoch": 0.47714761880713097,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00027184466019417475,
      "loss": 0.8826,
      "step": 1181
    },
    {
      "epoch": 0.4775516388061209,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00027182038834951453,
      "loss": 0.7676,
      "step": 1182
    },
    {
      "epoch": 0.4779556588051109,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00027179611650485436,
      "loss": 0.8612,
      "step": 1183
    },
    {
      "epoch": 0.4783596788041008,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00027177184466019413,
      "loss": 0.7875,
      "step": 1184
    },
    {
      "epoch": 0.4787636988030908,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00027174757281553396,
      "loss": 0.8312,
      "step": 1185
    },
    {
      "epoch": 0.4791677188020807,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002717233009708738,
      "loss": 0.7684,
      "step": 1186
    },
    {
      "epoch": 0.4795717388010707,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00027169902912621357,
      "loss": 0.931,
      "step": 1187
    },
    {
      "epoch": 0.4799757588000606,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002716747572815534,
      "loss": 0.9964,
      "step": 1188
    },
    {
      "epoch": 0.4803797787990505,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002716504854368932,
      "loss": 0.899,
      "step": 1189
    },
    {
      "epoch": 0.4807837987980405,
      "grad_norm": 0.49609375,
      "learning_rate": 0.000271626213592233,
      "loss": 0.8272,
      "step": 1190
    },
    {
      "epoch": 0.48118781879703043,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002716019417475728,
      "loss": 0.9005,
      "step": 1191
    },
    {
      "epoch": 0.4815918387960204,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002715776699029126,
      "loss": 0.8782,
      "step": 1192
    },
    {
      "epoch": 0.48199585879501033,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002715533980582524,
      "loss": 1.1729,
      "step": 1193
    },
    {
      "epoch": 0.4823998787940003,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002715291262135922,
      "loss": 0.8858,
      "step": 1194
    },
    {
      "epoch": 0.48280389879299024,
      "grad_norm": 0.5390625,
      "learning_rate": 0.000271504854368932,
      "loss": 0.8897,
      "step": 1195
    },
    {
      "epoch": 0.4832079187919802,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002714805825242718,
      "loss": 0.9079,
      "step": 1196
    },
    {
      "epoch": 0.48361193879097014,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00027145631067961165,
      "loss": 0.9274,
      "step": 1197
    },
    {
      "epoch": 0.4840159587899601,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002714320388349514,
      "loss": 0.958,
      "step": 1198
    },
    {
      "epoch": 0.48441997878895005,
      "grad_norm": 1.765625,
      "learning_rate": 0.00027140776699029125,
      "loss": 0.9068,
      "step": 1199
    },
    {
      "epoch": 0.48482399878794,
      "grad_norm": 0.65234375,
      "learning_rate": 0.000271383495145631,
      "loss": 0.8989,
      "step": 1200
    },
    {
      "epoch": 0.48522801878692995,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00027135922330097086,
      "loss": 0.848,
      "step": 1201
    },
    {
      "epoch": 0.48563203878591993,
      "grad_norm": 0.78125,
      "learning_rate": 0.0002713349514563107,
      "loss": 0.7974,
      "step": 1202
    },
    {
      "epoch": 0.48603605878490985,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00027131067961165046,
      "loss": 0.8219,
      "step": 1203
    },
    {
      "epoch": 0.4864400787838998,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00027128640776699024,
      "loss": 0.8716,
      "step": 1204
    },
    {
      "epoch": 0.48684409878288976,
      "grad_norm": 0.59375,
      "learning_rate": 0.00027126213592233006,
      "loss": 0.8425,
      "step": 1205
    },
    {
      "epoch": 0.4872481187818797,
      "grad_norm": 0.625,
      "learning_rate": 0.0002712378640776699,
      "loss": 0.9162,
      "step": 1206
    },
    {
      "epoch": 0.48765213878086966,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00027121359223300967,
      "loss": 0.797,
      "step": 1207
    },
    {
      "epoch": 0.4880561587798596,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002711893203883495,
      "loss": 0.8058,
      "step": 1208
    },
    {
      "epoch": 0.48846017877884956,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002711650485436893,
      "loss": 0.8699,
      "step": 1209
    },
    {
      "epoch": 0.4888641987778395,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002711407766990291,
      "loss": 0.9559,
      "step": 1210
    },
    {
      "epoch": 0.48926821877682947,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00027111650485436893,
      "loss": 1.0552,
      "step": 1211
    },
    {
      "epoch": 0.4896722387758194,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002710922330097087,
      "loss": 0.8338,
      "step": 1212
    },
    {
      "epoch": 0.49007625877480937,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00027106796116504854,
      "loss": 0.8045,
      "step": 1213
    },
    {
      "epoch": 0.4904802787737993,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002710436893203883,
      "loss": 0.8167,
      "step": 1214
    },
    {
      "epoch": 0.4908842987727893,
      "grad_norm": 0.859375,
      "learning_rate": 0.00027101941747572814,
      "loss": 0.9096,
      "step": 1215
    },
    {
      "epoch": 0.4912883187717792,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0002709951456310679,
      "loss": 0.9511,
      "step": 1216
    },
    {
      "epoch": 0.4916923387707692,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027097087378640775,
      "loss": 0.9182,
      "step": 1217
    },
    {
      "epoch": 0.4920963587697591,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002709466019417476,
      "loss": 0.9703,
      "step": 1218
    },
    {
      "epoch": 0.4925003787687491,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00027092233009708735,
      "loss": 0.9121,
      "step": 1219
    },
    {
      "epoch": 0.492904398767739,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00027089805825242713,
      "loss": 0.8757,
      "step": 1220
    },
    {
      "epoch": 0.49330841876672893,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00027087378640776696,
      "loss": 0.8899,
      "step": 1221
    },
    {
      "epoch": 0.4937124387657189,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002708495145631068,
      "loss": 0.9543,
      "step": 1222
    },
    {
      "epoch": 0.49411645876470883,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00027082524271844656,
      "loss": 0.817,
      "step": 1223
    },
    {
      "epoch": 0.4945204787636988,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002708009708737864,
      "loss": 0.8611,
      "step": 1224
    },
    {
      "epoch": 0.49492449876268874,
      "grad_norm": 0.7421875,
      "learning_rate": 0.00027077669902912617,
      "loss": 0.9017,
      "step": 1225
    },
    {
      "epoch": 0.4953285187616787,
      "grad_norm": 0.55078125,
      "learning_rate": 0.000270752427184466,
      "loss": 0.9115,
      "step": 1226
    },
    {
      "epoch": 0.49573253876066864,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002707281553398058,
      "loss": 0.8844,
      "step": 1227
    },
    {
      "epoch": 0.4961365587596586,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002707038834951456,
      "loss": 0.8466,
      "step": 1228
    },
    {
      "epoch": 0.49654057875864854,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00027067961165048543,
      "loss": 0.8262,
      "step": 1229
    },
    {
      "epoch": 0.4969445987576385,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002706553398058252,
      "loss": 0.8224,
      "step": 1230
    },
    {
      "epoch": 0.49734861875662845,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00027063106796116503,
      "loss": 0.888,
      "step": 1231
    },
    {
      "epoch": 0.4977526387556184,
      "grad_norm": 0.578125,
      "learning_rate": 0.00027060679611650486,
      "loss": 0.8551,
      "step": 1232
    },
    {
      "epoch": 0.49815665875460835,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00027058252427184464,
      "loss": 0.9137,
      "step": 1233
    },
    {
      "epoch": 0.49856067875359833,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002705582524271844,
      "loss": 0.8991,
      "step": 1234
    },
    {
      "epoch": 0.49896469875258825,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00027053398058252424,
      "loss": 0.8508,
      "step": 1235
    },
    {
      "epoch": 0.4993687187515782,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002705097087378641,
      "loss": 0.8386,
      "step": 1236
    },
    {
      "epoch": 0.49977273875056816,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00027048543689320385,
      "loss": 0.9948,
      "step": 1237
    },
    {
      "epoch": 0.5001767587495581,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002704611650485437,
      "loss": 0.7951,
      "step": 1238
    },
    {
      "epoch": 0.500580778748548,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00027043689320388345,
      "loss": 0.9541,
      "step": 1239
    },
    {
      "epoch": 0.500984798747538,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002704126213592233,
      "loss": 1.0335,
      "step": 1240
    },
    {
      "epoch": 0.501388818746528,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002703883495145631,
      "loss": 0.9389,
      "step": 1241
    },
    {
      "epoch": 0.5017928387455179,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002703640776699029,
      "loss": 0.9152,
      "step": 1242
    },
    {
      "epoch": 0.5021968587445078,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002703398058252427,
      "loss": 0.9164,
      "step": 1243
    },
    {
      "epoch": 0.5026008787434978,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002703155339805825,
      "loss": 0.8994,
      "step": 1244
    },
    {
      "epoch": 0.5030048987424878,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00027029126213592227,
      "loss": 0.8968,
      "step": 1245
    },
    {
      "epoch": 0.5034089187414778,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002702669902912621,
      "loss": 0.8585,
      "step": 1246
    },
    {
      "epoch": 0.5038129387404676,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002702427184466019,
      "loss": 0.9829,
      "step": 1247
    },
    {
      "epoch": 0.5042169587394576,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00027021844660194176,
      "loss": 0.9664,
      "step": 1248
    },
    {
      "epoch": 0.5046209787384476,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00027019417475728153,
      "loss": 0.9245,
      "step": 1249
    },
    {
      "epoch": 0.5050249987374376,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002701699029126213,
      "loss": 0.925,
      "step": 1250
    },
    {
      "epoch": 0.5054290187364274,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027014563106796114,
      "loss": 0.9181,
      "step": 1251
    },
    {
      "epoch": 0.5058330387354174,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00027012135922330096,
      "loss": 0.8261,
      "step": 1252
    },
    {
      "epoch": 0.5062370587344074,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00027009708737864074,
      "loss": 0.8602,
      "step": 1253
    },
    {
      "epoch": 0.5066410787333973,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00027007281553398057,
      "loss": 0.9508,
      "step": 1254
    },
    {
      "epoch": 0.5070450987323872,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00027004854368932034,
      "loss": 0.8274,
      "step": 1255
    },
    {
      "epoch": 0.5074491187313772,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002700242718446602,
      "loss": 0.8809,
      "step": 1256
    },
    {
      "epoch": 0.5078531387303672,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00027,
      "loss": 0.8377,
      "step": 1257
    },
    {
      "epoch": 0.5082571587293571,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002699757281553398,
      "loss": 0.9392,
      "step": 1258
    },
    {
      "epoch": 0.508661178728347,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002699514563106796,
      "loss": 0.9637,
      "step": 1259
    },
    {
      "epoch": 0.509065198727337,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002699271844660194,
      "loss": 0.8862,
      "step": 1260
    },
    {
      "epoch": 0.509469218726327,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002699029126213592,
      "loss": 1.0409,
      "step": 1261
    },
    {
      "epoch": 0.5098732387253169,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00026987864077669904,
      "loss": 1.0121,
      "step": 1262
    },
    {
      "epoch": 0.5102772587243068,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002698543689320388,
      "loss": 0.8254,
      "step": 1263
    },
    {
      "epoch": 0.5106812787232968,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002698300970873786,
      "loss": 0.8587,
      "step": 1264
    },
    {
      "epoch": 0.5110852987222868,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002698058252427184,
      "loss": 0.8594,
      "step": 1265
    },
    {
      "epoch": 0.5114893187212767,
      "grad_norm": 10.375,
      "learning_rate": 0.00026978155339805825,
      "loss": 0.9726,
      "step": 1266
    },
    {
      "epoch": 0.5118933387202667,
      "grad_norm": 0.51171875,
      "learning_rate": 0.000269757281553398,
      "loss": 0.8971,
      "step": 1267
    },
    {
      "epoch": 0.5122973587192566,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00026973300970873786,
      "loss": 0.9351,
      "step": 1268
    },
    {
      "epoch": 0.5127013787182465,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00026970873786407763,
      "loss": 0.8054,
      "step": 1269
    },
    {
      "epoch": 0.5131053987172365,
      "grad_norm": 0.546875,
      "learning_rate": 0.00026968446601941746,
      "loss": 0.8099,
      "step": 1270
    },
    {
      "epoch": 0.5135094187162265,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00026966019417475724,
      "loss": 0.9183,
      "step": 1271
    },
    {
      "epoch": 0.5139134387152164,
      "grad_norm": 0.609375,
      "learning_rate": 0.00026963592233009707,
      "loss": 0.8274,
      "step": 1272
    },
    {
      "epoch": 0.5143174587142063,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002696116504854369,
      "loss": 0.9206,
      "step": 1273
    },
    {
      "epoch": 0.5147214787131963,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00026958737864077667,
      "loss": 0.7776,
      "step": 1274
    },
    {
      "epoch": 0.5151254987121863,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00026956310679611645,
      "loss": 0.8768,
      "step": 1275
    },
    {
      "epoch": 0.5155295187111762,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002695388349514563,
      "loss": 0.8698,
      "step": 1276
    },
    {
      "epoch": 0.5159335387101661,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002695145631067961,
      "loss": 0.8156,
      "step": 1277
    },
    {
      "epoch": 0.5163375587091561,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002694902912621359,
      "loss": 0.8356,
      "step": 1278
    },
    {
      "epoch": 0.5167415787081461,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002694660194174757,
      "loss": 0.8301,
      "step": 1279
    },
    {
      "epoch": 0.5171455987071361,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002694417475728155,
      "loss": 0.8422,
      "step": 1280
    },
    {
      "epoch": 0.5175496187061259,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002694174757281553,
      "loss": 0.8379,
      "step": 1281
    },
    {
      "epoch": 0.5179536387051159,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00026939320388349514,
      "loss": 1.0175,
      "step": 1282
    },
    {
      "epoch": 0.5183576587041059,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002693689320388349,
      "loss": 0.9491,
      "step": 1283
    },
    {
      "epoch": 0.5187616787030958,
      "grad_norm": 0.59375,
      "learning_rate": 0.00026934466019417475,
      "loss": 0.8644,
      "step": 1284
    },
    {
      "epoch": 0.5191656987020857,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002693203883495145,
      "loss": 0.7552,
      "step": 1285
    },
    {
      "epoch": 0.5195697187010757,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00026929611650485435,
      "loss": 0.9852,
      "step": 1286
    },
    {
      "epoch": 0.5199737387000657,
      "grad_norm": 0.470703125,
      "learning_rate": 0.0002692718446601942,
      "loss": 0.7786,
      "step": 1287
    },
    {
      "epoch": 0.5203777586990556,
      "grad_norm": 0.625,
      "learning_rate": 0.00026924757281553396,
      "loss": 0.8616,
      "step": 1288
    },
    {
      "epoch": 0.5207817786980455,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002692233009708738,
      "loss": 0.8604,
      "step": 1289
    },
    {
      "epoch": 0.5211857986970355,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00026919902912621356,
      "loss": 0.8128,
      "step": 1290
    },
    {
      "epoch": 0.5215898186960255,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002691747572815534,
      "loss": 0.9263,
      "step": 1291
    },
    {
      "epoch": 0.5219938386950154,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00026915048543689317,
      "loss": 0.7916,
      "step": 1292
    },
    {
      "epoch": 0.5223978586940053,
      "grad_norm": 0.6171875,
      "learning_rate": 0.000269126213592233,
      "loss": 0.9226,
      "step": 1293
    },
    {
      "epoch": 0.5228018786929953,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00026910194174757277,
      "loss": 0.829,
      "step": 1294
    },
    {
      "epoch": 0.5232058986919853,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002690776699029126,
      "loss": 0.9897,
      "step": 1295
    },
    {
      "epoch": 0.5236099186909752,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002690533980582524,
      "loss": 0.9497,
      "step": 1296
    },
    {
      "epoch": 0.5240139386899652,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002690291262135922,
      "loss": 0.9457,
      "step": 1297
    },
    {
      "epoch": 0.5244179586889551,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00026900485436893203,
      "loss": 0.9793,
      "step": 1298
    },
    {
      "epoch": 0.524821978687945,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002689805825242718,
      "loss": 0.9198,
      "step": 1299
    },
    {
      "epoch": 0.525225998686935,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00026895631067961164,
      "loss": 1.003,
      "step": 1300
    },
    {
      "epoch": 0.525630018685925,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002689320388349514,
      "loss": 0.8591,
      "step": 1301
    },
    {
      "epoch": 0.5260340386849149,
      "grad_norm": 0.53125,
      "learning_rate": 0.00026890776699029124,
      "loss": 0.7966,
      "step": 1302
    },
    {
      "epoch": 0.5264380586839048,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002688834951456311,
      "loss": 0.8801,
      "step": 1303
    },
    {
      "epoch": 0.5268420786828948,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00026885922330097085,
      "loss": 0.9625,
      "step": 1304
    },
    {
      "epoch": 0.5272460986818848,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002688349514563106,
      "loss": 0.8616,
      "step": 1305
    },
    {
      "epoch": 0.5276501186808747,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00026881067961165045,
      "loss": 0.9248,
      "step": 1306
    },
    {
      "epoch": 0.5280541386798646,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002687864077669903,
      "loss": 0.9521,
      "step": 1307
    },
    {
      "epoch": 0.5284581586788546,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00026876213592233006,
      "loss": 1.0446,
      "step": 1308
    },
    {
      "epoch": 0.5288621786778446,
      "grad_norm": 0.4609375,
      "learning_rate": 0.0002687378640776699,
      "loss": 0.7983,
      "step": 1309
    },
    {
      "epoch": 0.5292661986768346,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00026871359223300966,
      "loss": 0.7752,
      "step": 1310
    },
    {
      "epoch": 0.5296702186758244,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002686893203883495,
      "loss": 0.977,
      "step": 1311
    },
    {
      "epoch": 0.5300742386748144,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002686650485436893,
      "loss": 0.8583,
      "step": 1312
    },
    {
      "epoch": 0.5304782586738044,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002686407766990291,
      "loss": 0.7913,
      "step": 1313
    },
    {
      "epoch": 0.5308822786727942,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002686165048543689,
      "loss": 0.8825,
      "step": 1314
    },
    {
      "epoch": 0.5312862986717842,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002685922330097087,
      "loss": 0.9315,
      "step": 1315
    },
    {
      "epoch": 0.5316903186707742,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00026856796116504853,
      "loss": 0.8631,
      "step": 1316
    },
    {
      "epoch": 0.5320943386697642,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002685436893203883,
      "loss": 0.8694,
      "step": 1317
    },
    {
      "epoch": 0.5324983586687541,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00026851941747572814,
      "loss": 0.9226,
      "step": 1318
    },
    {
      "epoch": 0.532902378667744,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00026849514563106797,
      "loss": 0.9228,
      "step": 1319
    },
    {
      "epoch": 0.533306398666734,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00026847087378640774,
      "loss": 0.8604,
      "step": 1320
    },
    {
      "epoch": 0.533710418665724,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002684466019417475,
      "loss": 0.7949,
      "step": 1321
    },
    {
      "epoch": 0.5341144386647139,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00026842233009708735,
      "loss": 0.8495,
      "step": 1322
    },
    {
      "epoch": 0.5345184586637038,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002683980582524272,
      "loss": 0.7764,
      "step": 1323
    },
    {
      "epoch": 0.5349224786626938,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00026837378640776695,
      "loss": 0.875,
      "step": 1324
    },
    {
      "epoch": 0.5353264986616838,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002683495145631068,
      "loss": 0.8986,
      "step": 1325
    },
    {
      "epoch": 0.5357305186606737,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00026832524271844655,
      "loss": 0.8497,
      "step": 1326
    },
    {
      "epoch": 0.5361345386596637,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002683009708737864,
      "loss": 0.9331,
      "step": 1327
    },
    {
      "epoch": 0.5365385586586536,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002682766990291262,
      "loss": 0.8472,
      "step": 1328
    },
    {
      "epoch": 0.5369425786576435,
      "grad_norm": 0.46484375,
      "learning_rate": 0.000268252427184466,
      "loss": 0.7695,
      "step": 1329
    },
    {
      "epoch": 0.5373465986566335,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002682281553398058,
      "loss": 0.8567,
      "step": 1330
    },
    {
      "epoch": 0.5377506186556235,
      "grad_norm": 0.46484375,
      "learning_rate": 0.0002682038834951456,
      "loss": 0.6902,
      "step": 1331
    },
    {
      "epoch": 0.5381546386546134,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002681796116504854,
      "loss": 0.8895,
      "step": 1332
    },
    {
      "epoch": 0.5385586586536033,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00026815533980582525,
      "loss": 0.8503,
      "step": 1333
    },
    {
      "epoch": 0.5389626786525933,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00026813106796116503,
      "loss": 0.9392,
      "step": 1334
    },
    {
      "epoch": 0.5393666986515833,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002681067961165048,
      "loss": 0.8602,
      "step": 1335
    },
    {
      "epoch": 0.5397707186505732,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00026808252427184463,
      "loss": 0.79,
      "step": 1336
    },
    {
      "epoch": 0.5401747386495631,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00026805825242718446,
      "loss": 0.9084,
      "step": 1337
    },
    {
      "epoch": 0.5405787586485531,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00026803398058252424,
      "loss": 0.8115,
      "step": 1338
    },
    {
      "epoch": 0.5409827786475431,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00026800970873786407,
      "loss": 0.814,
      "step": 1339
    },
    {
      "epoch": 0.541386798646533,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00026798543689320384,
      "loss": 0.8673,
      "step": 1340
    },
    {
      "epoch": 0.5417908186455229,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00026796116504854367,
      "loss": 0.8535,
      "step": 1341
    },
    {
      "epoch": 0.5421948386445129,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002679368932038835,
      "loss": 0.8312,
      "step": 1342
    },
    {
      "epoch": 0.5425988586435029,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002679126213592233,
      "loss": 0.8686,
      "step": 1343
    },
    {
      "epoch": 0.5430028786424929,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002678883495145631,
      "loss": 0.8561,
      "step": 1344
    },
    {
      "epoch": 0.5434068986414827,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002678640776699029,
      "loss": 0.811,
      "step": 1345
    },
    {
      "epoch": 0.5438109186404727,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00026783980582524266,
      "loss": 0.7897,
      "step": 1346
    },
    {
      "epoch": 0.5442149386394627,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002678155339805825,
      "loss": 0.7833,
      "step": 1347
    },
    {
      "epoch": 0.5446189586384526,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002677912621359223,
      "loss": 0.9201,
      "step": 1348
    },
    {
      "epoch": 0.5450229786374425,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00026776699029126214,
      "loss": 0.9207,
      "step": 1349
    },
    {
      "epoch": 0.5454269986364325,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002677427184466019,
      "loss": 0.8278,
      "step": 1350
    },
    {
      "epoch": 0.5458310186354225,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002677184466019417,
      "loss": 0.9118,
      "step": 1351
    },
    {
      "epoch": 0.5462350386344124,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002676941747572815,
      "loss": 0.9001,
      "step": 1352
    },
    {
      "epoch": 0.5466390586334023,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00026766990291262135,
      "loss": 0.9134,
      "step": 1353
    },
    {
      "epoch": 0.5470430786323923,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00026764563106796113,
      "loss": 0.7841,
      "step": 1354
    },
    {
      "epoch": 0.5474470986313823,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00026762135922330096,
      "loss": 0.8856,
      "step": 1355
    },
    {
      "epoch": 0.5478511186303722,
      "grad_norm": 0.609375,
      "learning_rate": 0.00026759708737864073,
      "loss": 0.8935,
      "step": 1356
    },
    {
      "epoch": 0.5482551386293621,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00026757281553398056,
      "loss": 0.915,
      "step": 1357
    },
    {
      "epoch": 0.5486591586283521,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002675485436893204,
      "loss": 0.8841,
      "step": 1358
    },
    {
      "epoch": 0.5490631786273421,
      "grad_norm": 0.609375,
      "learning_rate": 0.00026752427184466017,
      "loss": 0.8439,
      "step": 1359
    },
    {
      "epoch": 0.549467198626332,
      "grad_norm": 0.5,
      "learning_rate": 0.0002675,
      "loss": 0.7601,
      "step": 1360
    },
    {
      "epoch": 0.549871218625322,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00026747572815533977,
      "loss": 0.8422,
      "step": 1361
    },
    {
      "epoch": 0.5502752386243119,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002674514563106796,
      "loss": 0.8543,
      "step": 1362
    },
    {
      "epoch": 0.5506792586233018,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00026742718446601943,
      "loss": 0.959,
      "step": 1363
    },
    {
      "epoch": 0.5510832786222918,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002674029126213592,
      "loss": 0.9374,
      "step": 1364
    },
    {
      "epoch": 0.5514872986212818,
      "grad_norm": 0.5625,
      "learning_rate": 0.000267378640776699,
      "loss": 0.9523,
      "step": 1365
    },
    {
      "epoch": 0.5518913186202717,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002673543689320388,
      "loss": 0.8168,
      "step": 1366
    },
    {
      "epoch": 0.5522953386192616,
      "grad_norm": 0.5625,
      "learning_rate": 0.00026733009708737864,
      "loss": 0.8726,
      "step": 1367
    },
    {
      "epoch": 0.5526993586182516,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002673058252427184,
      "loss": 0.8816,
      "step": 1368
    },
    {
      "epoch": 0.5531033786172416,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00026728155339805825,
      "loss": 0.9373,
      "step": 1369
    },
    {
      "epoch": 0.5535073986162315,
      "grad_norm": 0.52734375,
      "learning_rate": 0.000267257281553398,
      "loss": 0.8561,
      "step": 1370
    },
    {
      "epoch": 0.5539114186152214,
      "grad_norm": 0.578125,
      "learning_rate": 0.00026723300970873785,
      "loss": 0.7618,
      "step": 1371
    },
    {
      "epoch": 0.5543154386142114,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002672087378640776,
      "loss": 0.9182,
      "step": 1372
    },
    {
      "epoch": 0.5547194586132014,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00026718446601941745,
      "loss": 0.7475,
      "step": 1373
    },
    {
      "epoch": 0.5551234786121914,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002671601941747573,
      "loss": 0.8739,
      "step": 1374
    },
    {
      "epoch": 0.5555274986111812,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00026713592233009706,
      "loss": 0.8771,
      "step": 1375
    },
    {
      "epoch": 0.5559315186101712,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00026711165048543683,
      "loss": 0.9402,
      "step": 1376
    },
    {
      "epoch": 0.5563355386091612,
      "grad_norm": 0.625,
      "learning_rate": 0.00026708737864077666,
      "loss": 0.9041,
      "step": 1377
    },
    {
      "epoch": 0.556739558608151,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002670631067961165,
      "loss": 0.9188,
      "step": 1378
    },
    {
      "epoch": 0.557143578607141,
      "grad_norm": 0.5625,
      "learning_rate": 0.00026703883495145627,
      "loss": 0.8953,
      "step": 1379
    },
    {
      "epoch": 0.557547598606131,
      "grad_norm": 23.125,
      "learning_rate": 0.0002670145631067961,
      "loss": 1.0533,
      "step": 1380
    },
    {
      "epoch": 0.557951618605121,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002669902912621359,
      "loss": 0.9172,
      "step": 1381
    },
    {
      "epoch": 0.5583556386041109,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002669660194174757,
      "loss": 0.9318,
      "step": 1382
    },
    {
      "epoch": 0.5587596586031008,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00026694174757281553,
      "loss": 0.9607,
      "step": 1383
    },
    {
      "epoch": 0.5591636786020908,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002669174757281553,
      "loss": 0.857,
      "step": 1384
    },
    {
      "epoch": 0.5595676986010808,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00026689320388349514,
      "loss": 0.9261,
      "step": 1385
    },
    {
      "epoch": 0.5599717186000707,
      "grad_norm": 0.43359375,
      "learning_rate": 0.0002668689320388349,
      "loss": 0.798,
      "step": 1386
    },
    {
      "epoch": 0.5603757385990606,
      "grad_norm": 0.59375,
      "learning_rate": 0.00026684466019417474,
      "loss": 0.8823,
      "step": 1387
    },
    {
      "epoch": 0.5607797585980506,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00026682038834951457,
      "loss": 0.8425,
      "step": 1388
    },
    {
      "epoch": 0.5611837785970406,
      "grad_norm": 0.53125,
      "learning_rate": 0.00026679611650485435,
      "loss": 0.7664,
      "step": 1389
    },
    {
      "epoch": 0.5615877985960305,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002667718446601942,
      "loss": 0.8678,
      "step": 1390
    },
    {
      "epoch": 0.5619918185950205,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00026674757281553395,
      "loss": 0.8871,
      "step": 1391
    },
    {
      "epoch": 0.5623958385940104,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002667233009708738,
      "loss": 0.7607,
      "step": 1392
    },
    {
      "epoch": 0.5627998585930003,
      "grad_norm": 0.578125,
      "learning_rate": 0.00026669902912621356,
      "loss": 0.8138,
      "step": 1393
    },
    {
      "epoch": 0.5632038785919903,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002666747572815534,
      "loss": 0.8115,
      "step": 1394
    },
    {
      "epoch": 0.5636078985909803,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00026665048543689316,
      "loss": 0.8145,
      "step": 1395
    },
    {
      "epoch": 0.5640119185899702,
      "grad_norm": 0.55078125,
      "learning_rate": 0.000266626213592233,
      "loss": 0.88,
      "step": 1396
    },
    {
      "epoch": 0.5644159385889601,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00026660194174757276,
      "loss": 0.9771,
      "step": 1397
    },
    {
      "epoch": 0.5648199585879501,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002665776699029126,
      "loss": 0.8715,
      "step": 1398
    },
    {
      "epoch": 0.5652239785869401,
      "grad_norm": 0.75,
      "learning_rate": 0.0002665533980582524,
      "loss": 0.8296,
      "step": 1399
    },
    {
      "epoch": 0.56562799858593,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002665291262135922,
      "loss": 0.9253,
      "step": 1400
    },
    {
      "epoch": 0.5660320185849199,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00026650485436893203,
      "loss": 0.827,
      "step": 1401
    },
    {
      "epoch": 0.5664360385839099,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002664805825242718,
      "loss": 0.8737,
      "step": 1402
    },
    {
      "epoch": 0.5668400585828999,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00026645631067961163,
      "loss": 0.9696,
      "step": 1403
    },
    {
      "epoch": 0.5672440785818899,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00026643203883495146,
      "loss": 0.8544,
      "step": 1404
    },
    {
      "epoch": 0.5676480985808797,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00026640776699029124,
      "loss": 0.828,
      "step": 1405
    },
    {
      "epoch": 0.5680521185798697,
      "grad_norm": 0.609375,
      "learning_rate": 0.000266383495145631,
      "loss": 0.9082,
      "step": 1406
    },
    {
      "epoch": 0.5684561385788597,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00026635922330097084,
      "loss": 0.7719,
      "step": 1407
    },
    {
      "epoch": 0.5688601585778496,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00026633495145631067,
      "loss": 0.7672,
      "step": 1408
    },
    {
      "epoch": 0.5692641785768395,
      "grad_norm": 0.6875,
      "learning_rate": 0.00026631067961165045,
      "loss": 0.9306,
      "step": 1409
    },
    {
      "epoch": 0.5696681985758295,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002662864077669903,
      "loss": 0.784,
      "step": 1410
    },
    {
      "epoch": 0.5700722185748195,
      "grad_norm": 0.53125,
      "learning_rate": 0.00026626213592233005,
      "loss": 0.8627,
      "step": 1411
    },
    {
      "epoch": 0.5704762385738094,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002662378640776699,
      "loss": 0.8744,
      "step": 1412
    },
    {
      "epoch": 0.5708802585727993,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002662135922330097,
      "loss": 0.8406,
      "step": 1413
    },
    {
      "epoch": 0.5712842785717893,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002661893203883495,
      "loss": 0.9038,
      "step": 1414
    },
    {
      "epoch": 0.5716882985707793,
      "grad_norm": 0.45703125,
      "learning_rate": 0.0002661650485436893,
      "loss": 0.7531,
      "step": 1415
    },
    {
      "epoch": 0.5720923185697692,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002661407766990291,
      "loss": 0.8826,
      "step": 1416
    },
    {
      "epoch": 0.5724963385687591,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002661165048543689,
      "loss": 0.8811,
      "step": 1417
    },
    {
      "epoch": 0.5729003585677491,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002660922330097087,
      "loss": 0.8921,
      "step": 1418
    },
    {
      "epoch": 0.5733043785667391,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002660679611650485,
      "loss": 0.846,
      "step": 1419
    },
    {
      "epoch": 0.573708398565729,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00026604368932038835,
      "loss": 0.7617,
      "step": 1420
    },
    {
      "epoch": 0.574112418564719,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00026601941747572813,
      "loss": 1.0166,
      "step": 1421
    },
    {
      "epoch": 0.5745164385637089,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002659951456310679,
      "loss": 0.697,
      "step": 1422
    },
    {
      "epoch": 0.5749204585626988,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00026597087378640773,
      "loss": 0.9225,
      "step": 1423
    },
    {
      "epoch": 0.5753244785616888,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00026594660194174756,
      "loss": 0.7972,
      "step": 1424
    },
    {
      "epoch": 0.5757284985606788,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00026592233009708734,
      "loss": 0.8962,
      "step": 1425
    },
    {
      "epoch": 0.5761325185596687,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00026589805825242717,
      "loss": 0.8348,
      "step": 1426
    },
    {
      "epoch": 0.5765365385586586,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00026587378640776694,
      "loss": 0.8568,
      "step": 1427
    },
    {
      "epoch": 0.5769405585576486,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00026584951456310677,
      "loss": 0.8632,
      "step": 1428
    },
    {
      "epoch": 0.5773445785566386,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002658252427184466,
      "loss": 0.954,
      "step": 1429
    },
    {
      "epoch": 0.5777485985556285,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002658009708737864,
      "loss": 0.8269,
      "step": 1430
    },
    {
      "epoch": 0.5781526185546184,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002657766990291262,
      "loss": 0.8599,
      "step": 1431
    },
    {
      "epoch": 0.5785566385536084,
      "grad_norm": 0.58984375,
      "learning_rate": 0.000265752427184466,
      "loss": 0.8896,
      "step": 1432
    },
    {
      "epoch": 0.5789606585525984,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002657281553398058,
      "loss": 0.9318,
      "step": 1433
    },
    {
      "epoch": 0.5793646785515884,
      "grad_norm": 0.5625,
      "learning_rate": 0.00026570388349514564,
      "loss": 0.741,
      "step": 1434
    },
    {
      "epoch": 0.5797686985505782,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002656796116504854,
      "loss": 0.8059,
      "step": 1435
    },
    {
      "epoch": 0.5801727185495682,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002656553398058252,
      "loss": 0.9362,
      "step": 1436
    },
    {
      "epoch": 0.5805767385485582,
      "grad_norm": 0.64453125,
      "learning_rate": 0.000265631067961165,
      "loss": 0.8147,
      "step": 1437
    },
    {
      "epoch": 0.580980758547548,
      "grad_norm": 0.53125,
      "learning_rate": 0.00026560679611650485,
      "loss": 0.7813,
      "step": 1438
    },
    {
      "epoch": 0.581384778546538,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002655825242718446,
      "loss": 0.8497,
      "step": 1439
    },
    {
      "epoch": 0.581788798545528,
      "grad_norm": 0.546875,
      "learning_rate": 0.00026555825242718446,
      "loss": 0.8207,
      "step": 1440
    },
    {
      "epoch": 0.582192818544518,
      "grad_norm": 0.578125,
      "learning_rate": 0.00026553398058252423,
      "loss": 0.8307,
      "step": 1441
    },
    {
      "epoch": 0.5825968385435079,
      "grad_norm": 0.625,
      "learning_rate": 0.00026550970873786406,
      "loss": 0.9556,
      "step": 1442
    },
    {
      "epoch": 0.5830008585424978,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002654854368932039,
      "loss": 0.8701,
      "step": 1443
    },
    {
      "epoch": 0.5834048785414878,
      "grad_norm": 0.671875,
      "learning_rate": 0.00026546116504854366,
      "loss": 0.928,
      "step": 1444
    },
    {
      "epoch": 0.5838088985404778,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002654368932038835,
      "loss": 0.7628,
      "step": 1445
    },
    {
      "epoch": 0.5842129185394677,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00026541262135922327,
      "loss": 0.8048,
      "step": 1446
    },
    {
      "epoch": 0.5846169385384576,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00026538834951456304,
      "loss": 0.8467,
      "step": 1447
    },
    {
      "epoch": 0.5850209585374476,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002653640776699029,
      "loss": 0.8846,
      "step": 1448
    },
    {
      "epoch": 0.5854249785364376,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002653398058252427,
      "loss": 0.8429,
      "step": 1449
    },
    {
      "epoch": 0.5858289985354275,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00026531553398058253,
      "loss": 0.8839,
      "step": 1450
    },
    {
      "epoch": 0.5862330185344174,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002652912621359223,
      "loss": 0.8515,
      "step": 1451
    },
    {
      "epoch": 0.5866370385334074,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002652669902912621,
      "loss": 0.8253,
      "step": 1452
    },
    {
      "epoch": 0.5870410585323974,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002652427184466019,
      "loss": 0.8056,
      "step": 1453
    },
    {
      "epoch": 0.5874450785313873,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00026521844660194174,
      "loss": 0.8781,
      "step": 1454
    },
    {
      "epoch": 0.5878490985303773,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002651941747572815,
      "loss": 0.8379,
      "step": 1455
    },
    {
      "epoch": 0.5882531185293672,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00026516990291262135,
      "loss": 0.8203,
      "step": 1456
    },
    {
      "epoch": 0.5886571385283571,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002651456310679611,
      "loss": 0.8279,
      "step": 1457
    },
    {
      "epoch": 0.5890611585273471,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00026512135922330095,
      "loss": 0.9737,
      "step": 1458
    },
    {
      "epoch": 0.5894651785263371,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002650970873786408,
      "loss": 0.919,
      "step": 1459
    },
    {
      "epoch": 0.589869198525327,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00026507281553398056,
      "loss": 0.8825,
      "step": 1460
    },
    {
      "epoch": 0.5902732185243169,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002650485436893204,
      "loss": 0.928,
      "step": 1461
    },
    {
      "epoch": 0.5906772385233069,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00026502427184466016,
      "loss": 0.9006,
      "step": 1462
    },
    {
      "epoch": 0.5910812585222969,
      "grad_norm": 0.640625,
      "learning_rate": 0.000265,
      "loss": 0.9638,
      "step": 1463
    },
    {
      "epoch": 0.5914852785212869,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002649757281553398,
      "loss": 0.8346,
      "step": 1464
    },
    {
      "epoch": 0.5918892985202767,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002649514563106796,
      "loss": 0.9346,
      "step": 1465
    },
    {
      "epoch": 0.5922933185192667,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00026492718446601937,
      "loss": 0.8417,
      "step": 1466
    },
    {
      "epoch": 0.5926973385182567,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002649029126213592,
      "loss": 0.9153,
      "step": 1467
    },
    {
      "epoch": 0.5931013585172467,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00026487864077669903,
      "loss": 0.9231,
      "step": 1468
    },
    {
      "epoch": 0.5935053785162365,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002648543689320388,
      "loss": 0.7981,
      "step": 1469
    },
    {
      "epoch": 0.5939093985152265,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00026483009708737863,
      "loss": 0.8825,
      "step": 1470
    },
    {
      "epoch": 0.5943134185142165,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002648058252427184,
      "loss": 0.8667,
      "step": 1471
    },
    {
      "epoch": 0.5947174385132064,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00026478155339805824,
      "loss": 0.744,
      "step": 1472
    },
    {
      "epoch": 0.5951214585121963,
      "grad_norm": 0.62109375,
      "learning_rate": 0.000264757281553398,
      "loss": 0.9815,
      "step": 1473
    },
    {
      "epoch": 0.5955254785111863,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00026473300970873784,
      "loss": 0.8079,
      "step": 1474
    },
    {
      "epoch": 0.5959294985101763,
      "grad_norm": 0.484375,
      "learning_rate": 0.00026470873786407767,
      "loss": 0.7513,
      "step": 1475
    },
    {
      "epoch": 0.5963335185091662,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00026468446601941745,
      "loss": 0.8679,
      "step": 1476
    },
    {
      "epoch": 0.5967375385081561,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002646601941747572,
      "loss": 0.8251,
      "step": 1477
    },
    {
      "epoch": 0.5971415585071461,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00026463592233009705,
      "loss": 0.7255,
      "step": 1478
    },
    {
      "epoch": 0.5975455785061361,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002646116504854369,
      "loss": 0.895,
      "step": 1479
    },
    {
      "epoch": 0.597949598505126,
      "grad_norm": 0.578125,
      "learning_rate": 0.00026458737864077666,
      "loss": 0.8851,
      "step": 1480
    },
    {
      "epoch": 0.598353618504116,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002645631067961165,
      "loss": 0.8974,
      "step": 1481
    },
    {
      "epoch": 0.5987576385031059,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00026453883495145626,
      "loss": 0.9873,
      "step": 1482
    },
    {
      "epoch": 0.5991616585020959,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002645145631067961,
      "loss": 0.882,
      "step": 1483
    },
    {
      "epoch": 0.5995656785010858,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002644902912621359,
      "loss": 0.9511,
      "step": 1484
    },
    {
      "epoch": 0.5999696985000758,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002644660194174757,
      "loss": 0.9369,
      "step": 1485
    },
    {
      "epoch": 0.6003737184990657,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002644417475728155,
      "loss": 0.8879,
      "step": 1486
    },
    {
      "epoch": 0.6007777384980556,
      "grad_norm": 0.8125,
      "learning_rate": 0.0002644174757281553,
      "loss": 1.0533,
      "step": 1487
    },
    {
      "epoch": 0.6011817584970456,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00026439320388349513,
      "loss": 0.891,
      "step": 1488
    },
    {
      "epoch": 0.6015857784960356,
      "grad_norm": 0.546875,
      "learning_rate": 0.00026436893203883496,
      "loss": 0.836,
      "step": 1489
    },
    {
      "epoch": 0.6019897984950255,
      "grad_norm": 0.609375,
      "learning_rate": 0.00026434466019417473,
      "loss": 0.881,
      "step": 1490
    },
    {
      "epoch": 0.6023938184940154,
      "grad_norm": 0.53125,
      "learning_rate": 0.00026432038834951456,
      "loss": 0.8601,
      "step": 1491
    },
    {
      "epoch": 0.6027978384930054,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00026429611650485434,
      "loss": 0.8605,
      "step": 1492
    },
    {
      "epoch": 0.6032018584919954,
      "grad_norm": 0.609375,
      "learning_rate": 0.00026427184466019417,
      "loss": 0.9349,
      "step": 1493
    },
    {
      "epoch": 0.6036058784909853,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00026424757281553394,
      "loss": 0.8687,
      "step": 1494
    },
    {
      "epoch": 0.6040098984899752,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002642233009708738,
      "loss": 0.7745,
      "step": 1495
    },
    {
      "epoch": 0.6044139184889652,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00026419902912621355,
      "loss": 0.8211,
      "step": 1496
    },
    {
      "epoch": 0.6048179384879552,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002641747572815534,
      "loss": 0.9613,
      "step": 1497
    },
    {
      "epoch": 0.6052219584869452,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00026415048543689315,
      "loss": 0.8203,
      "step": 1498
    },
    {
      "epoch": 0.605625978485935,
      "grad_norm": 0.609375,
      "learning_rate": 0.000264126213592233,
      "loss": 1.0009,
      "step": 1499
    },
    {
      "epoch": 0.606029998484925,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002641019417475728,
      "loss": 0.8339,
      "step": 1500
    },
    {
      "epoch": 0.606434018483915,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002640776699029126,
      "loss": 0.8334,
      "step": 1501
    },
    {
      "epoch": 0.6068380384829049,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0002640533980582524,
      "loss": 1.026,
      "step": 1502
    },
    {
      "epoch": 0.6072420584818948,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002640291262135922,
      "loss": 0.8851,
      "step": 1503
    },
    {
      "epoch": 0.6076460784808848,
      "grad_norm": 0.5859375,
      "learning_rate": 0.000264004854368932,
      "loss": 0.9841,
      "step": 1504
    },
    {
      "epoch": 0.6080500984798748,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00026398058252427185,
      "loss": 0.8624,
      "step": 1505
    },
    {
      "epoch": 0.6084541184788647,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002639563106796116,
      "loss": 0.8132,
      "step": 1506
    },
    {
      "epoch": 0.6088581384778546,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002639320388349514,
      "loss": 0.7483,
      "step": 1507
    },
    {
      "epoch": 0.6092621584768446,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00026390776699029123,
      "loss": 0.8643,
      "step": 1508
    },
    {
      "epoch": 0.6096661784758346,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00026388349514563106,
      "loss": 0.8543,
      "step": 1509
    },
    {
      "epoch": 0.6100701984748245,
      "grad_norm": 0.484375,
      "learning_rate": 0.00026385922330097084,
      "loss": 0.8106,
      "step": 1510
    },
    {
      "epoch": 0.6104742184738144,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00026383495145631067,
      "loss": 0.8402,
      "step": 1511
    },
    {
      "epoch": 0.6108782384728044,
      "grad_norm": 0.515625,
      "learning_rate": 0.00026381067961165044,
      "loss": 0.8676,
      "step": 1512
    },
    {
      "epoch": 0.6112822584717944,
      "grad_norm": 0.703125,
      "learning_rate": 0.00026378640776699027,
      "loss": 0.8976,
      "step": 1513
    },
    {
      "epoch": 0.6116862784707843,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002637621359223301,
      "loss": 0.9,
      "step": 1514
    },
    {
      "epoch": 0.6120902984697743,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002637378640776699,
      "loss": 0.8517,
      "step": 1515
    },
    {
      "epoch": 0.6124943184687642,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002637135922330097,
      "loss": 0.852,
      "step": 1516
    },
    {
      "epoch": 0.6128983384677541,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002636893203883495,
      "loss": 0.8336,
      "step": 1517
    },
    {
      "epoch": 0.6133023584667441,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002636650485436893,
      "loss": 0.8887,
      "step": 1518
    },
    {
      "epoch": 0.6137063784657341,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002636407766990291,
      "loss": 0.9086,
      "step": 1519
    },
    {
      "epoch": 0.614110398464724,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002636165048543689,
      "loss": 0.7106,
      "step": 1520
    },
    {
      "epoch": 0.6145144184637139,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00026359223300970874,
      "loss": 0.8399,
      "step": 1521
    },
    {
      "epoch": 0.6149184384627039,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002635679611650485,
      "loss": 0.7809,
      "step": 1522
    },
    {
      "epoch": 0.6153224584616939,
      "grad_norm": 0.4609375,
      "learning_rate": 0.0002635436893203883,
      "loss": 0.7927,
      "step": 1523
    },
    {
      "epoch": 0.6157264784606838,
      "grad_norm": 0.796875,
      "learning_rate": 0.0002635194174757281,
      "loss": 0.8558,
      "step": 1524
    },
    {
      "epoch": 0.6161304984596737,
      "grad_norm": 0.640625,
      "learning_rate": 0.00026349514563106795,
      "loss": 0.9081,
      "step": 1525
    },
    {
      "epoch": 0.6165345184586637,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00026347087378640773,
      "loss": 0.7776,
      "step": 1526
    },
    {
      "epoch": 0.6169385384576537,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00026344660194174756,
      "loss": 0.7591,
      "step": 1527
    },
    {
      "epoch": 0.6173425584566437,
      "grad_norm": 0.53125,
      "learning_rate": 0.00026342233009708733,
      "loss": 0.7919,
      "step": 1528
    },
    {
      "epoch": 0.6177465784556335,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00026339805825242716,
      "loss": 0.8048,
      "step": 1529
    },
    {
      "epoch": 0.6181505984546235,
      "grad_norm": 0.59375,
      "learning_rate": 0.000263373786407767,
      "loss": 0.8535,
      "step": 1530
    },
    {
      "epoch": 0.6185546184536135,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00026334951456310677,
      "loss": 0.8246,
      "step": 1531
    },
    {
      "epoch": 0.6189586384526033,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002633252427184466,
      "loss": 0.8646,
      "step": 1532
    },
    {
      "epoch": 0.6193626584515933,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00026330097087378637,
      "loss": 0.7301,
      "step": 1533
    },
    {
      "epoch": 0.6197666784505833,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002632766990291262,
      "loss": 0.796,
      "step": 1534
    },
    {
      "epoch": 0.6201706984495733,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00026325242718446603,
      "loss": 0.8621,
      "step": 1535
    },
    {
      "epoch": 0.6205747184485632,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002632281553398058,
      "loss": 0.7405,
      "step": 1536
    },
    {
      "epoch": 0.6209787384475531,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002632038834951456,
      "loss": 0.7882,
      "step": 1537
    },
    {
      "epoch": 0.6213827584465431,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002631796116504854,
      "loss": 0.9358,
      "step": 1538
    },
    {
      "epoch": 0.6217867784455331,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00026315533980582524,
      "loss": 0.7745,
      "step": 1539
    },
    {
      "epoch": 0.622190798444523,
      "grad_norm": 0.625,
      "learning_rate": 0.000263131067961165,
      "loss": 0.8686,
      "step": 1540
    },
    {
      "epoch": 0.6225948184435129,
      "grad_norm": 0.65625,
      "learning_rate": 0.00026310679611650484,
      "loss": 0.8572,
      "step": 1541
    },
    {
      "epoch": 0.6229988384425029,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002630825242718446,
      "loss": 0.8297,
      "step": 1542
    },
    {
      "epoch": 0.6234028584414929,
      "grad_norm": 0.625,
      "learning_rate": 0.00026305825242718445,
      "loss": 0.7796,
      "step": 1543
    },
    {
      "epoch": 0.6238068784404828,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002630339805825242,
      "loss": 0.9628,
      "step": 1544
    },
    {
      "epoch": 0.6242108984394727,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00026300970873786405,
      "loss": 0.7887,
      "step": 1545
    },
    {
      "epoch": 0.6246149184384627,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002629854368932039,
      "loss": 0.9062,
      "step": 1546
    },
    {
      "epoch": 0.6250189384374527,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00026296116504854366,
      "loss": 0.8268,
      "step": 1547
    },
    {
      "epoch": 0.6254229584364426,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00026293689320388343,
      "loss": 0.9628,
      "step": 1548
    },
    {
      "epoch": 0.6258269784354326,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00026291262135922326,
      "loss": 0.8598,
      "step": 1549
    },
    {
      "epoch": 0.6262309984344225,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002628883495145631,
      "loss": 0.8596,
      "step": 1550
    },
    {
      "epoch": 0.6266350184334124,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0002628640776699029,
      "loss": 0.8437,
      "step": 1551
    },
    {
      "epoch": 0.6270390384324024,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002628398058252427,
      "loss": 0.9297,
      "step": 1552
    },
    {
      "epoch": 0.6274430584313924,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00026281553398058247,
      "loss": 0.8867,
      "step": 1553
    },
    {
      "epoch": 0.6278470784303823,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002627912621359223,
      "loss": 0.92,
      "step": 1554
    },
    {
      "epoch": 0.6282510984293722,
      "grad_norm": 0.640625,
      "learning_rate": 0.00026276699029126213,
      "loss": 0.9713,
      "step": 1555
    },
    {
      "epoch": 0.6286551184283622,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002627427184466019,
      "loss": 0.8656,
      "step": 1556
    },
    {
      "epoch": 0.6290591384273522,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00026271844660194174,
      "loss": 0.8678,
      "step": 1557
    },
    {
      "epoch": 0.6294631584263422,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002626941747572815,
      "loss": 0.8887,
      "step": 1558
    },
    {
      "epoch": 0.629867178425332,
      "grad_norm": 0.71875,
      "learning_rate": 0.00026266990291262134,
      "loss": 0.872,
      "step": 1559
    },
    {
      "epoch": 0.630271198424322,
      "grad_norm": 0.609375,
      "learning_rate": 0.00026264563106796117,
      "loss": 0.7638,
      "step": 1560
    },
    {
      "epoch": 0.630675218423312,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00026262135922330094,
      "loss": 0.7935,
      "step": 1561
    },
    {
      "epoch": 0.631079238422302,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002625970873786408,
      "loss": 0.8592,
      "step": 1562
    },
    {
      "epoch": 0.6314832584212918,
      "grad_norm": 0.59375,
      "learning_rate": 0.00026257281553398055,
      "loss": 0.8783,
      "step": 1563
    },
    {
      "epoch": 0.6318872784202818,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002625485436893204,
      "loss": 0.8388,
      "step": 1564
    },
    {
      "epoch": 0.6322912984192718,
      "grad_norm": 10.75,
      "learning_rate": 0.0002625242718446602,
      "loss": 0.8766,
      "step": 1565
    },
    {
      "epoch": 0.6326953184182617,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0002625,
      "loss": 0.8473,
      "step": 1566
    },
    {
      "epoch": 0.6330993384172516,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00026247572815533976,
      "loss": 0.7857,
      "step": 1567
    },
    {
      "epoch": 0.6335033584162416,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002624514563106796,
      "loss": 0.8694,
      "step": 1568
    },
    {
      "epoch": 0.6339073784152316,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002624271844660194,
      "loss": 0.8763,
      "step": 1569
    },
    {
      "epoch": 0.6343113984142215,
      "grad_norm": 0.478515625,
      "learning_rate": 0.0002624029126213592,
      "loss": 0.8392,
      "step": 1570
    },
    {
      "epoch": 0.6347154184132114,
      "grad_norm": 0.5859375,
      "learning_rate": 0.000262378640776699,
      "loss": 0.7515,
      "step": 1571
    },
    {
      "epoch": 0.6351194384122014,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002623543689320388,
      "loss": 0.8578,
      "step": 1572
    },
    {
      "epoch": 0.6355234584111914,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00026233009708737863,
      "loss": 0.7757,
      "step": 1573
    },
    {
      "epoch": 0.6359274784101813,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002623058252427184,
      "loss": 0.8749,
      "step": 1574
    },
    {
      "epoch": 0.6363314984091712,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00026228155339805823,
      "loss": 0.7762,
      "step": 1575
    },
    {
      "epoch": 0.6367355184081612,
      "grad_norm": 0.484375,
      "learning_rate": 0.00026225728155339806,
      "loss": 0.801,
      "step": 1576
    },
    {
      "epoch": 0.6371395384071512,
      "grad_norm": 0.515625,
      "learning_rate": 0.00026223300970873784,
      "loss": 0.8372,
      "step": 1577
    },
    {
      "epoch": 0.6375435584061411,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002622087378640776,
      "loss": 0.8079,
      "step": 1578
    },
    {
      "epoch": 0.637947578405131,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00026218446601941744,
      "loss": 0.9032,
      "step": 1579
    },
    {
      "epoch": 0.638351598404121,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00026216019417475727,
      "loss": 0.9409,
      "step": 1580
    },
    {
      "epoch": 0.6387556184031109,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00026213592233009705,
      "loss": 0.9317,
      "step": 1581
    },
    {
      "epoch": 0.6391596384021009,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002621116504854369,
      "loss": 0.7918,
      "step": 1582
    },
    {
      "epoch": 0.6395636584010909,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00026208737864077665,
      "loss": 0.7524,
      "step": 1583
    },
    {
      "epoch": 0.6399676784000808,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002620631067961165,
      "loss": 0.8121,
      "step": 1584
    },
    {
      "epoch": 0.6403716983990707,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0002620388349514563,
      "loss": 0.9001,
      "step": 1585
    },
    {
      "epoch": 0.6407757183980607,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002620145631067961,
      "loss": 0.7823,
      "step": 1586
    },
    {
      "epoch": 0.6411797383970507,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002619902912621359,
      "loss": 0.8913,
      "step": 1587
    },
    {
      "epoch": 0.6415837583960406,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002619660194174757,
      "loss": 0.6801,
      "step": 1588
    },
    {
      "epoch": 0.6419877783950305,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002619417475728155,
      "loss": 0.9737,
      "step": 1589
    },
    {
      "epoch": 0.6423917983940205,
      "grad_norm": 0.76171875,
      "learning_rate": 0.00026191747572815535,
      "loss": 0.9765,
      "step": 1590
    },
    {
      "epoch": 0.6427958183930105,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002618932038834951,
      "loss": 0.8534,
      "step": 1591
    },
    {
      "epoch": 0.6431998383920005,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00026186893203883495,
      "loss": 0.8907,
      "step": 1592
    },
    {
      "epoch": 0.6436038583909903,
      "grad_norm": 0.53125,
      "learning_rate": 0.00026184466019417473,
      "loss": 0.8464,
      "step": 1593
    },
    {
      "epoch": 0.6440078783899803,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00026182038834951456,
      "loss": 0.8277,
      "step": 1594
    },
    {
      "epoch": 0.6444118983889703,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00026179611650485433,
      "loss": 0.8341,
      "step": 1595
    },
    {
      "epoch": 0.6448159183879602,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00026177184466019416,
      "loss": 0.8328,
      "step": 1596
    },
    {
      "epoch": 0.6452199383869501,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00026174757281553394,
      "loss": 0.8189,
      "step": 1597
    },
    {
      "epoch": 0.6456239583859401,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00026172330097087377,
      "loss": 0.8785,
      "step": 1598
    },
    {
      "epoch": 0.6460279783849301,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00026169902912621354,
      "loss": 0.992,
      "step": 1599
    },
    {
      "epoch": 0.64643199838392,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00026167475728155337,
      "loss": 0.8591,
      "step": 1600
    },
    {
      "epoch": 0.6468360183829099,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002616504854368932,
      "loss": 0.8045,
      "step": 1601
    },
    {
      "epoch": 0.6472400383818999,
      "grad_norm": 0.5625,
      "learning_rate": 0.000261626213592233,
      "loss": 0.7515,
      "step": 1602
    },
    {
      "epoch": 0.6476440583808899,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0002616019417475728,
      "loss": 0.8424,
      "step": 1603
    },
    {
      "epoch": 0.6480480783798798,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002615776699029126,
      "loss": 0.8114,
      "step": 1604
    },
    {
      "epoch": 0.6484520983788697,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002615533980582524,
      "loss": 0.8419,
      "step": 1605
    },
    {
      "epoch": 0.6488561183778597,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00026152912621359224,
      "loss": 0.8351,
      "step": 1606
    },
    {
      "epoch": 0.6492601383768497,
      "grad_norm": 0.490234375,
      "learning_rate": 0.000261504854368932,
      "loss": 0.7582,
      "step": 1607
    },
    {
      "epoch": 0.6496641583758396,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002614805825242718,
      "loss": 0.8756,
      "step": 1608
    },
    {
      "epoch": 0.6500681783748296,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002614563106796116,
      "loss": 0.8338,
      "step": 1609
    },
    {
      "epoch": 0.6504721983738195,
      "grad_norm": 0.5625,
      "learning_rate": 0.00026143203883495145,
      "loss": 0.8204,
      "step": 1610
    },
    {
      "epoch": 0.6508762183728094,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002614077669902912,
      "loss": 0.8597,
      "step": 1611
    },
    {
      "epoch": 0.6512802383717994,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00026138349514563105,
      "loss": 0.8046,
      "step": 1612
    },
    {
      "epoch": 0.6516842583707894,
      "grad_norm": 0.5625,
      "learning_rate": 0.00026135922330097083,
      "loss": 0.8544,
      "step": 1613
    },
    {
      "epoch": 0.6520882783697793,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00026133495145631066,
      "loss": 1.023,
      "step": 1614
    },
    {
      "epoch": 0.6524922983687692,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002613106796116505,
      "loss": 0.7778,
      "step": 1615
    },
    {
      "epoch": 0.6528963183677592,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026128640776699026,
      "loss": 0.7586,
      "step": 1616
    },
    {
      "epoch": 0.6533003383667492,
      "grad_norm": 0.4609375,
      "learning_rate": 0.0002612621359223301,
      "loss": 0.7047,
      "step": 1617
    },
    {
      "epoch": 0.6537043583657391,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00026123786407766987,
      "loss": 0.9178,
      "step": 1618
    },
    {
      "epoch": 0.654108378364729,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0002612135922330097,
      "loss": 0.9086,
      "step": 1619
    },
    {
      "epoch": 0.654512398363719,
      "grad_norm": 0.578125,
      "learning_rate": 0.00026118932038834947,
      "loss": 0.9267,
      "step": 1620
    },
    {
      "epoch": 0.654916418362709,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002611650485436893,
      "loss": 0.883,
      "step": 1621
    },
    {
      "epoch": 0.655320438361699,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00026114077669902913,
      "loss": 0.8088,
      "step": 1622
    },
    {
      "epoch": 0.6557244583606888,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002611165048543689,
      "loss": 0.9362,
      "step": 1623
    },
    {
      "epoch": 0.6561284783596788,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002610922330097087,
      "loss": 0.82,
      "step": 1624
    },
    {
      "epoch": 0.6565324983586688,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002610679611650485,
      "loss": 0.753,
      "step": 1625
    },
    {
      "epoch": 0.6569365183576586,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00026104368932038834,
      "loss": 0.7307,
      "step": 1626
    },
    {
      "epoch": 0.6573405383566486,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002610194174757281,
      "loss": 0.8488,
      "step": 1627
    },
    {
      "epoch": 0.6577445583556386,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00026099514563106795,
      "loss": 0.8737,
      "step": 1628
    },
    {
      "epoch": 0.6581485783546286,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002609708737864077,
      "loss": 0.8815,
      "step": 1629
    },
    {
      "epoch": 0.6585525983536185,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00026094660194174755,
      "loss": 0.8425,
      "step": 1630
    },
    {
      "epoch": 0.6589566183526084,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002609223300970874,
      "loss": 0.8403,
      "step": 1631
    },
    {
      "epoch": 0.6593606383515984,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00026089805825242716,
      "loss": 0.8852,
      "step": 1632
    },
    {
      "epoch": 0.6597646583505884,
      "grad_norm": 0.51953125,
      "learning_rate": 0.000260873786407767,
      "loss": 0.7646,
      "step": 1633
    },
    {
      "epoch": 0.6601686783495783,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00026084951456310676,
      "loss": 0.8933,
      "step": 1634
    },
    {
      "epoch": 0.6605726983485682,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002608252427184466,
      "loss": 0.8618,
      "step": 1635
    },
    {
      "epoch": 0.6609767183475582,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002608009708737864,
      "loss": 0.7939,
      "step": 1636
    },
    {
      "epoch": 0.6613807383465482,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002607766990291262,
      "loss": 0.9402,
      "step": 1637
    },
    {
      "epoch": 0.6617847583455381,
      "grad_norm": 0.5625,
      "learning_rate": 0.00026075242718446597,
      "loss": 0.9615,
      "step": 1638
    },
    {
      "epoch": 0.662188778344528,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0002607281553398058,
      "loss": 0.9714,
      "step": 1639
    },
    {
      "epoch": 0.662592798343518,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00026070388349514563,
      "loss": 0.7713,
      "step": 1640
    },
    {
      "epoch": 0.6629968183425079,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002606796116504854,
      "loss": 0.957,
      "step": 1641
    },
    {
      "epoch": 0.6634008383414979,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00026065533980582523,
      "loss": 0.9017,
      "step": 1642
    },
    {
      "epoch": 0.6638048583404879,
      "grad_norm": 0.546875,
      "learning_rate": 0.000260631067961165,
      "loss": 0.9436,
      "step": 1643
    },
    {
      "epoch": 0.6642088783394778,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00026060679611650484,
      "loss": 0.7444,
      "step": 1644
    },
    {
      "epoch": 0.6646128983384677,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002605825242718446,
      "loss": 0.8836,
      "step": 1645
    },
    {
      "epoch": 0.6650169183374577,
      "grad_norm": 0.65625,
      "learning_rate": 0.00026055825242718444,
      "loss": 0.9279,
      "step": 1646
    },
    {
      "epoch": 0.6654209383364477,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00026053398058252427,
      "loss": 0.992,
      "step": 1647
    },
    {
      "epoch": 0.6658249583354376,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00026050970873786405,
      "loss": 0.9085,
      "step": 1648
    },
    {
      "epoch": 0.6662289783344275,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002604854368932038,
      "loss": 0.8339,
      "step": 1649
    },
    {
      "epoch": 0.6666329983334175,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00026046116504854365,
      "loss": 0.8012,
      "step": 1650
    },
    {
      "epoch": 0.6670370183324075,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002604368932038835,
      "loss": 0.7868,
      "step": 1651
    },
    {
      "epoch": 0.6674410383313975,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002604126213592233,
      "loss": 0.9254,
      "step": 1652
    },
    {
      "epoch": 0.6678450583303873,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002603883495145631,
      "loss": 0.8265,
      "step": 1653
    },
    {
      "epoch": 0.6682490783293773,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00026036407766990286,
      "loss": 0.7228,
      "step": 1654
    },
    {
      "epoch": 0.6686530983283673,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002603398058252427,
      "loss": 0.8516,
      "step": 1655
    },
    {
      "epoch": 0.6690571183273573,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002603155339805825,
      "loss": 0.8268,
      "step": 1656
    },
    {
      "epoch": 0.6694611383263471,
      "grad_norm": 0.435546875,
      "learning_rate": 0.0002602912621359223,
      "loss": 0.8074,
      "step": 1657
    },
    {
      "epoch": 0.6698651583253371,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002602669902912621,
      "loss": 0.8148,
      "step": 1658
    },
    {
      "epoch": 0.6702691783243271,
      "grad_norm": 0.48046875,
      "learning_rate": 0.0002602427184466019,
      "loss": 0.8223,
      "step": 1659
    },
    {
      "epoch": 0.670673198323317,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00026021844660194173,
      "loss": 0.753,
      "step": 1660
    },
    {
      "epoch": 0.6710772183223069,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00026019417475728156,
      "loss": 0.7832,
      "step": 1661
    },
    {
      "epoch": 0.6714812383212969,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00026016990291262133,
      "loss": 0.9772,
      "step": 1662
    },
    {
      "epoch": 0.6718852583202869,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00026014563106796116,
      "loss": 0.9529,
      "step": 1663
    },
    {
      "epoch": 0.6722892783192768,
      "grad_norm": 0.53125,
      "learning_rate": 0.00026012135922330094,
      "loss": 0.8717,
      "step": 1664
    },
    {
      "epoch": 0.6726932983182667,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00026009708737864077,
      "loss": 0.9791,
      "step": 1665
    },
    {
      "epoch": 0.6730973183172567,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002600728155339806,
      "loss": 0.8087,
      "step": 1666
    },
    {
      "epoch": 0.6735013383162467,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00026004854368932037,
      "loss": 0.8142,
      "step": 1667
    },
    {
      "epoch": 0.6739053583152366,
      "grad_norm": 0.515625,
      "learning_rate": 0.00026002427184466015,
      "loss": 0.7557,
      "step": 1668
    },
    {
      "epoch": 0.6743093783142265,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00026,
      "loss": 0.9216,
      "step": 1669
    },
    {
      "epoch": 0.6747133983132165,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002599757281553398,
      "loss": 0.7794,
      "step": 1670
    },
    {
      "epoch": 0.6751174183122065,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002599514563106796,
      "loss": 0.9087,
      "step": 1671
    },
    {
      "epoch": 0.6755214383111964,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002599271844660194,
      "loss": 0.8142,
      "step": 1672
    },
    {
      "epoch": 0.6759254583101864,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002599029126213592,
      "loss": 0.9093,
      "step": 1673
    },
    {
      "epoch": 0.6763294783091763,
      "grad_norm": 0.53125,
      "learning_rate": 0.000259878640776699,
      "loss": 0.7984,
      "step": 1674
    },
    {
      "epoch": 0.6767334983081662,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002598543689320388,
      "loss": 0.8211,
      "step": 1675
    },
    {
      "epoch": 0.6771375183071562,
      "grad_norm": 0.46484375,
      "learning_rate": 0.0002598300970873786,
      "loss": 0.7566,
      "step": 1676
    },
    {
      "epoch": 0.6775415383061462,
      "grad_norm": 0.5625,
      "learning_rate": 0.00025980582524271845,
      "loss": 0.9492,
      "step": 1677
    },
    {
      "epoch": 0.6779455583051361,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002597815533980582,
      "loss": 0.8231,
      "step": 1678
    },
    {
      "epoch": 0.678349578304126,
      "grad_norm": 0.5390625,
      "learning_rate": 0.000259757281553398,
      "loss": 0.8303,
      "step": 1679
    },
    {
      "epoch": 0.678753598303116,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00025973300970873783,
      "loss": 0.8797,
      "step": 1680
    },
    {
      "epoch": 0.679157618302106,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00025970873786407766,
      "loss": 0.7776,
      "step": 1681
    },
    {
      "epoch": 0.679561638301096,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00025968446601941743,
      "loss": 0.7801,
      "step": 1682
    },
    {
      "epoch": 0.6799656583000858,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00025966019417475726,
      "loss": 0.8728,
      "step": 1683
    },
    {
      "epoch": 0.6803696782990758,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00025963592233009704,
      "loss": 0.8154,
      "step": 1684
    },
    {
      "epoch": 0.6807736982980658,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00025961165048543687,
      "loss": 0.9054,
      "step": 1685
    },
    {
      "epoch": 0.6811777182970558,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002595873786407767,
      "loss": 0.9182,
      "step": 1686
    },
    {
      "epoch": 0.6815817382960456,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002595631067961165,
      "loss": 0.8584,
      "step": 1687
    },
    {
      "epoch": 0.6819857582950356,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002595388349514563,
      "loss": 0.8433,
      "step": 1688
    },
    {
      "epoch": 0.6823897782940256,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002595145631067961,
      "loss": 0.8373,
      "step": 1689
    },
    {
      "epoch": 0.6827937982930155,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002594902912621359,
      "loss": 0.8246,
      "step": 1690
    },
    {
      "epoch": 0.6831978182920054,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00025946601941747574,
      "loss": 0.8005,
      "step": 1691
    },
    {
      "epoch": 0.6836018382909954,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002594417475728155,
      "loss": 0.873,
      "step": 1692
    },
    {
      "epoch": 0.6840058582899854,
      "grad_norm": 0.546875,
      "learning_rate": 0.00025941747572815534,
      "loss": 0.8595,
      "step": 1693
    },
    {
      "epoch": 0.6844098782889753,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002593932038834951,
      "loss": 0.8171,
      "step": 1694
    },
    {
      "epoch": 0.6848138982879652,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00025936893203883495,
      "loss": 0.7807,
      "step": 1695
    },
    {
      "epoch": 0.6852179182869552,
      "grad_norm": 0.48046875,
      "learning_rate": 0.0002593446601941747,
      "loss": 0.8237,
      "step": 1696
    },
    {
      "epoch": 0.6856219382859452,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00025932038834951455,
      "loss": 0.8707,
      "step": 1697
    },
    {
      "epoch": 0.6860259582849351,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002592961165048543,
      "loss": 0.8332,
      "step": 1698
    },
    {
      "epoch": 0.686429978283925,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00025927184466019416,
      "loss": 0.7912,
      "step": 1699
    },
    {
      "epoch": 0.686833998282915,
      "grad_norm": 0.484375,
      "learning_rate": 0.00025924757281553393,
      "loss": 0.783,
      "step": 1700
    },
    {
      "epoch": 0.687238018281905,
      "grad_norm": 0.5625,
      "learning_rate": 0.00025922330097087376,
      "loss": 0.9004,
      "step": 1701
    },
    {
      "epoch": 0.6876420382808949,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002591990291262136,
      "loss": 1.0038,
      "step": 1702
    },
    {
      "epoch": 0.6880460582798849,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00025917475728155337,
      "loss": 0.8143,
      "step": 1703
    },
    {
      "epoch": 0.6884500782788748,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002591504854368932,
      "loss": 0.8283,
      "step": 1704
    },
    {
      "epoch": 0.6888540982778647,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00025912621359223297,
      "loss": 0.9488,
      "step": 1705
    },
    {
      "epoch": 0.6892581182768547,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002591019417475728,
      "loss": 0.9083,
      "step": 1706
    },
    {
      "epoch": 0.6896621382758447,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00025907766990291263,
      "loss": 0.813,
      "step": 1707
    },
    {
      "epoch": 0.6900661582748346,
      "grad_norm": 2.640625,
      "learning_rate": 0.0002590533980582524,
      "loss": 0.9248,
      "step": 1708
    },
    {
      "epoch": 0.6904701782738245,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002590291262135922,
      "loss": 0.9025,
      "step": 1709
    },
    {
      "epoch": 0.6908741982728145,
      "grad_norm": 0.52734375,
      "learning_rate": 0.000259004854368932,
      "loss": 0.7858,
      "step": 1710
    },
    {
      "epoch": 0.6912782182718045,
      "grad_norm": 0.578125,
      "learning_rate": 0.00025898058252427184,
      "loss": 0.8969,
      "step": 1711
    },
    {
      "epoch": 0.6916822382707944,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002589563106796116,
      "loss": 0.8747,
      "step": 1712
    },
    {
      "epoch": 0.6920862582697843,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00025893203883495144,
      "loss": 0.8751,
      "step": 1713
    },
    {
      "epoch": 0.6924902782687743,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002589077669902912,
      "loss": 0.779,
      "step": 1714
    },
    {
      "epoch": 0.6928942982677643,
      "grad_norm": 0.75,
      "learning_rate": 0.00025888349514563105,
      "loss": 0.8276,
      "step": 1715
    },
    {
      "epoch": 0.6932983182667543,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002588592233009709,
      "loss": 0.8251,
      "step": 1716
    },
    {
      "epoch": 0.6937023382657441,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00025883495145631065,
      "loss": 0.89,
      "step": 1717
    },
    {
      "epoch": 0.6941063582647341,
      "grad_norm": 0.671875,
      "learning_rate": 0.0002588106796116505,
      "loss": 0.9096,
      "step": 1718
    },
    {
      "epoch": 0.6945103782637241,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00025878640776699026,
      "loss": 0.9023,
      "step": 1719
    },
    {
      "epoch": 0.694914398262714,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002587621359223301,
      "loss": 0.803,
      "step": 1720
    },
    {
      "epoch": 0.6953184182617039,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00025873786407766986,
      "loss": 0.8224,
      "step": 1721
    },
    {
      "epoch": 0.6957224382606939,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0002587135922330097,
      "loss": 0.865,
      "step": 1722
    },
    {
      "epoch": 0.6961264582596839,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002586893203883495,
      "loss": 0.771,
      "step": 1723
    },
    {
      "epoch": 0.6965304782586738,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002586650485436893,
      "loss": 0.7785,
      "step": 1724
    },
    {
      "epoch": 0.6969344982576637,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00025864077669902907,
      "loss": 0.7762,
      "step": 1725
    },
    {
      "epoch": 0.6973385182566537,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002586165048543689,
      "loss": 0.852,
      "step": 1726
    },
    {
      "epoch": 0.6977425382556437,
      "grad_norm": 0.5,
      "learning_rate": 0.00025859223300970873,
      "loss": 0.8337,
      "step": 1727
    },
    {
      "epoch": 0.6981465582546336,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002585679611650485,
      "loss": 0.9023,
      "step": 1728
    },
    {
      "epoch": 0.6985505782536235,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00025854368932038833,
      "loss": 0.8488,
      "step": 1729
    },
    {
      "epoch": 0.6989545982526135,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002585194174757281,
      "loss": 0.7977,
      "step": 1730
    },
    {
      "epoch": 0.6993586182516035,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00025849514563106794,
      "loss": 0.7607,
      "step": 1731
    },
    {
      "epoch": 0.6997626382505934,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00025847087378640777,
      "loss": 1.0481,
      "step": 1732
    },
    {
      "epoch": 0.7001666582495834,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00025844660194174754,
      "loss": 0.7442,
      "step": 1733
    },
    {
      "epoch": 0.7005706782485733,
      "grad_norm": 0.5,
      "learning_rate": 0.0002584223300970874,
      "loss": 0.8008,
      "step": 1734
    },
    {
      "epoch": 0.7009746982475632,
      "grad_norm": 0.703125,
      "learning_rate": 0.00025839805825242715,
      "loss": 0.9379,
      "step": 1735
    },
    {
      "epoch": 0.7013787182465532,
      "grad_norm": 0.61328125,
      "learning_rate": 0.000258373786407767,
      "loss": 0.9058,
      "step": 1736
    },
    {
      "epoch": 0.7017827382455432,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002583495145631068,
      "loss": 0.9504,
      "step": 1737
    },
    {
      "epoch": 0.7021867582445331,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002583252427184466,
      "loss": 0.907,
      "step": 1738
    },
    {
      "epoch": 0.702590778243523,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00025830097087378636,
      "loss": 0.8134,
      "step": 1739
    },
    {
      "epoch": 0.702994798242513,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0002582766990291262,
      "loss": 0.9219,
      "step": 1740
    },
    {
      "epoch": 0.703398818241503,
      "grad_norm": 0.546875,
      "learning_rate": 0.000258252427184466,
      "loss": 0.8357,
      "step": 1741
    },
    {
      "epoch": 0.703802838240493,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002582281553398058,
      "loss": 0.8951,
      "step": 1742
    },
    {
      "epoch": 0.7042068582394828,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002582038834951456,
      "loss": 0.7786,
      "step": 1743
    },
    {
      "epoch": 0.7046108782384728,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002581796116504854,
      "loss": 0.7897,
      "step": 1744
    },
    {
      "epoch": 0.7050148982374628,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002581553398058252,
      "loss": 0.7769,
      "step": 1745
    },
    {
      "epoch": 0.7054189182364528,
      "grad_norm": 0.5234375,
      "learning_rate": 0.000258131067961165,
      "loss": 0.8079,
      "step": 1746
    },
    {
      "epoch": 0.7058229382354426,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00025810679611650483,
      "loss": 0.8809,
      "step": 1747
    },
    {
      "epoch": 0.7062269582344326,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00025808252427184466,
      "loss": 0.892,
      "step": 1748
    },
    {
      "epoch": 0.7066309782334226,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00025805825242718444,
      "loss": 0.9397,
      "step": 1749
    },
    {
      "epoch": 0.7070349982324126,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002580339805825242,
      "loss": 0.8077,
      "step": 1750
    },
    {
      "epoch": 0.7074390182314024,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00025800970873786404,
      "loss": 0.9855,
      "step": 1751
    },
    {
      "epoch": 0.7078430382303924,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00025798543689320387,
      "loss": 0.8312,
      "step": 1752
    },
    {
      "epoch": 0.7082470582293824,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002579611650485437,
      "loss": 0.8648,
      "step": 1753
    },
    {
      "epoch": 0.7086510782283723,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002579368932038835,
      "loss": 0.8785,
      "step": 1754
    },
    {
      "epoch": 0.7090550982273622,
      "grad_norm": 0.53125,
      "learning_rate": 0.00025791262135922325,
      "loss": 0.8303,
      "step": 1755
    },
    {
      "epoch": 0.7094591182263522,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002578883495145631,
      "loss": 0.8023,
      "step": 1756
    },
    {
      "epoch": 0.7098631382253422,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002578640776699029,
      "loss": 0.9063,
      "step": 1757
    },
    {
      "epoch": 0.7102671582243321,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002578398058252427,
      "loss": 0.8493,
      "step": 1758
    },
    {
      "epoch": 0.710671178223322,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002578155339805825,
      "loss": 0.9583,
      "step": 1759
    },
    {
      "epoch": 0.711075198222312,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002577912621359223,
      "loss": 0.8141,
      "step": 1760
    },
    {
      "epoch": 0.711479218221302,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002577669902912621,
      "loss": 0.9046,
      "step": 1761
    },
    {
      "epoch": 0.7118832382202919,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00025774271844660195,
      "loss": 0.8559,
      "step": 1762
    },
    {
      "epoch": 0.7122872582192818,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0002577184466019417,
      "loss": 0.961,
      "step": 1763
    },
    {
      "epoch": 0.7126912782182718,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00025769417475728155,
      "loss": 0.7478,
      "step": 1764
    },
    {
      "epoch": 0.7130952982172618,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00025766990291262133,
      "loss": 0.885,
      "step": 1765
    },
    {
      "epoch": 0.7134993182162517,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00025764563106796116,
      "loss": 0.8306,
      "step": 1766
    },
    {
      "epoch": 0.7139033382152417,
      "grad_norm": 0.58984375,
      "learning_rate": 0.000257621359223301,
      "loss": 0.9117,
      "step": 1767
    },
    {
      "epoch": 0.7143073582142316,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00025759708737864076,
      "loss": 0.7534,
      "step": 1768
    },
    {
      "epoch": 0.7147113782132215,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00025757281553398054,
      "loss": 0.8442,
      "step": 1769
    },
    {
      "epoch": 0.7151153982122115,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00025754854368932037,
      "loss": 0.7731,
      "step": 1770
    },
    {
      "epoch": 0.7155194182112015,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00025752427184466014,
      "loss": 0.8251,
      "step": 1771
    },
    {
      "epoch": 0.7159234382101914,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00025749999999999997,
      "loss": 0.8504,
      "step": 1772
    },
    {
      "epoch": 0.7163274582091813,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002574757281553398,
      "loss": 0.8677,
      "step": 1773
    },
    {
      "epoch": 0.7167314782081713,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002574514563106796,
      "loss": 0.8526,
      "step": 1774
    },
    {
      "epoch": 0.7171354982071613,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002574271844660194,
      "loss": 0.9193,
      "step": 1775
    },
    {
      "epoch": 0.7175395182061513,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002574029126213592,
      "loss": 0.8379,
      "step": 1776
    },
    {
      "epoch": 0.7179435382051411,
      "grad_norm": 0.484375,
      "learning_rate": 0.000257378640776699,
      "loss": 0.8221,
      "step": 1777
    },
    {
      "epoch": 0.7183475582041311,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00025735436893203884,
      "loss": 0.858,
      "step": 1778
    },
    {
      "epoch": 0.7187515782031211,
      "grad_norm": 0.470703125,
      "learning_rate": 0.0002573300970873786,
      "loss": 0.7306,
      "step": 1779
    },
    {
      "epoch": 0.7191555982021111,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002573058252427184,
      "loss": 1.0154,
      "step": 1780
    },
    {
      "epoch": 0.7195596182011009,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002572815533980582,
      "loss": 0.8555,
      "step": 1781
    },
    {
      "epoch": 0.7199636382000909,
      "grad_norm": 0.484375,
      "learning_rate": 0.00025725728155339805,
      "loss": 0.755,
      "step": 1782
    },
    {
      "epoch": 0.7203676581990809,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0002572330097087378,
      "loss": 1.0106,
      "step": 1783
    },
    {
      "epoch": 0.7207716781980708,
      "grad_norm": 0.5625,
      "learning_rate": 0.00025720873786407765,
      "loss": 0.8106,
      "step": 1784
    },
    {
      "epoch": 0.7211756981970607,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00025718446601941743,
      "loss": 0.8004,
      "step": 1785
    },
    {
      "epoch": 0.7215797181960507,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00025716019417475726,
      "loss": 0.8083,
      "step": 1786
    },
    {
      "epoch": 0.7219837381950407,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002571359223300971,
      "loss": 0.7499,
      "step": 1787
    },
    {
      "epoch": 0.7223877581940306,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00025711165048543686,
      "loss": 0.8362,
      "step": 1788
    },
    {
      "epoch": 0.7227917781930205,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002570873786407767,
      "loss": 0.866,
      "step": 1789
    },
    {
      "epoch": 0.7231957981920105,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00025706310679611647,
      "loss": 0.9669,
      "step": 1790
    },
    {
      "epoch": 0.7235998181910005,
      "grad_norm": 0.48046875,
      "learning_rate": 0.0002570388349514563,
      "loss": 0.748,
      "step": 1791
    },
    {
      "epoch": 0.7240038381899904,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002570145631067961,
      "loss": 0.9098,
      "step": 1792
    },
    {
      "epoch": 0.7244078581889803,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002569902912621359,
      "loss": 0.8389,
      "step": 1793
    },
    {
      "epoch": 0.7248118781879703,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00025696601941747573,
      "loss": 0.8068,
      "step": 1794
    },
    {
      "epoch": 0.7252158981869603,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002569417475728155,
      "loss": 0.9114,
      "step": 1795
    },
    {
      "epoch": 0.7256199181859502,
      "grad_norm": 0.73828125,
      "learning_rate": 0.00025691747572815534,
      "loss": 0.9136,
      "step": 1796
    },
    {
      "epoch": 0.7260239381849402,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002568932038834951,
      "loss": 0.8574,
      "step": 1797
    },
    {
      "epoch": 0.7264279581839301,
      "grad_norm": 0.546875,
      "learning_rate": 0.00025686893203883494,
      "loss": 0.8397,
      "step": 1798
    },
    {
      "epoch": 0.72683197818292,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002568446601941747,
      "loss": 0.8049,
      "step": 1799
    },
    {
      "epoch": 0.72723599818191,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00025682038834951454,
      "loss": 0.6389,
      "step": 1800
    },
    {
      "epoch": 0.7276400181809,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002567961165048543,
      "loss": 0.8799,
      "step": 1801
    },
    {
      "epoch": 0.7280440381798899,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00025677184466019415,
      "loss": 0.9025,
      "step": 1802
    },
    {
      "epoch": 0.7284480581788798,
      "grad_norm": 0.447265625,
      "learning_rate": 0.000256747572815534,
      "loss": 0.8093,
      "step": 1803
    },
    {
      "epoch": 0.7288520781778698,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00025672330097087375,
      "loss": 0.8179,
      "step": 1804
    },
    {
      "epoch": 0.7292560981768598,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002566990291262136,
      "loss": 0.688,
      "step": 1805
    },
    {
      "epoch": 0.7296601181758497,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00025667475728155336,
      "loss": 0.8309,
      "step": 1806
    },
    {
      "epoch": 0.7300641381748396,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002566504854368932,
      "loss": 0.775,
      "step": 1807
    },
    {
      "epoch": 0.7304681581738296,
      "grad_norm": 0.578125,
      "learning_rate": 0.000256626213592233,
      "loss": 0.8834,
      "step": 1808
    },
    {
      "epoch": 0.7308721781728196,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002566019417475728,
      "loss": 0.7341,
      "step": 1809
    },
    {
      "epoch": 0.7312761981718096,
      "grad_norm": 0.46875,
      "learning_rate": 0.00025657766990291257,
      "loss": 0.7763,
      "step": 1810
    },
    {
      "epoch": 0.7316802181707994,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002565533980582524,
      "loss": 0.8456,
      "step": 1811
    },
    {
      "epoch": 0.7320842381697894,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002565291262135922,
      "loss": 0.7771,
      "step": 1812
    },
    {
      "epoch": 0.7324882581687794,
      "grad_norm": 0.578125,
      "learning_rate": 0.000256504854368932,
      "loss": 0.8539,
      "step": 1813
    },
    {
      "epoch": 0.7328922781677693,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00025648058252427183,
      "loss": 0.8267,
      "step": 1814
    },
    {
      "epoch": 0.7332962981667592,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002564563106796116,
      "loss": 0.7828,
      "step": 1815
    },
    {
      "epoch": 0.7337003181657492,
      "grad_norm": 0.5625,
      "learning_rate": 0.00025643203883495144,
      "loss": 0.8447,
      "step": 1816
    },
    {
      "epoch": 0.7341043381647392,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00025640776699029127,
      "loss": 0.9017,
      "step": 1817
    },
    {
      "epoch": 0.7345083581637291,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00025638349514563104,
      "loss": 0.8866,
      "step": 1818
    },
    {
      "epoch": 0.734912378162719,
      "grad_norm": 0.59375,
      "learning_rate": 0.00025635922330097087,
      "loss": 0.8241,
      "step": 1819
    },
    {
      "epoch": 0.735316398161709,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00025633495145631065,
      "loss": 0.7949,
      "step": 1820
    },
    {
      "epoch": 0.735720418160699,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0002563106796116505,
      "loss": 1.0072,
      "step": 1821
    },
    {
      "epoch": 0.7361244381596889,
      "grad_norm": 0.53125,
      "learning_rate": 0.00025628640776699025,
      "loss": 0.8291,
      "step": 1822
    },
    {
      "epoch": 0.7365284581586788,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002562621359223301,
      "loss": 0.7817,
      "step": 1823
    },
    {
      "epoch": 0.7369324781576688,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002562378640776699,
      "loss": 0.8822,
      "step": 1824
    },
    {
      "epoch": 0.7373364981566588,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002562135922330097,
      "loss": 0.8385,
      "step": 1825
    },
    {
      "epoch": 0.7377405181556487,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00025618932038834946,
      "loss": 0.7055,
      "step": 1826
    },
    {
      "epoch": 0.7381445381546387,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002561650485436893,
      "loss": 0.7493,
      "step": 1827
    },
    {
      "epoch": 0.7385485581536286,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002561407766990291,
      "loss": 0.9298,
      "step": 1828
    },
    {
      "epoch": 0.7389525781526185,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0002561165048543689,
      "loss": 0.8167,
      "step": 1829
    },
    {
      "epoch": 0.7393565981516085,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002560922330097087,
      "loss": 1.0511,
      "step": 1830
    },
    {
      "epoch": 0.7397606181505985,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002560679611650485,
      "loss": 0.9132,
      "step": 1831
    },
    {
      "epoch": 0.7401646381495884,
      "grad_norm": 0.5625,
      "learning_rate": 0.00025604368932038833,
      "loss": 0.7909,
      "step": 1832
    },
    {
      "epoch": 0.7405686581485783,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00025601941747572816,
      "loss": 0.7946,
      "step": 1833
    },
    {
      "epoch": 0.7409726781475683,
      "grad_norm": 0.625,
      "learning_rate": 0.00025599514563106793,
      "loss": 0.8096,
      "step": 1834
    },
    {
      "epoch": 0.7413766981465583,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00025597087378640776,
      "loss": 0.8051,
      "step": 1835
    },
    {
      "epoch": 0.7417807181455482,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00025594660194174754,
      "loss": 0.826,
      "step": 1836
    },
    {
      "epoch": 0.7421847381445381,
      "grad_norm": 0.546875,
      "learning_rate": 0.00025592233009708737,
      "loss": 0.8319,
      "step": 1837
    },
    {
      "epoch": 0.7425887581435281,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002558980582524272,
      "loss": 0.7899,
      "step": 1838
    },
    {
      "epoch": 0.7429927781425181,
      "grad_norm": 0.5,
      "learning_rate": 0.00025587378640776697,
      "loss": 0.7917,
      "step": 1839
    },
    {
      "epoch": 0.743396798141508,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00025584951456310675,
      "loss": 0.9038,
      "step": 1840
    },
    {
      "epoch": 0.7438008181404979,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002558252427184466,
      "loss": 0.9313,
      "step": 1841
    },
    {
      "epoch": 0.7442048381394879,
      "grad_norm": 0.484375,
      "learning_rate": 0.0002558009708737864,
      "loss": 0.7767,
      "step": 1842
    },
    {
      "epoch": 0.7446088581384779,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002557766990291262,
      "loss": 0.8683,
      "step": 1843
    },
    {
      "epoch": 0.7450128781374679,
      "grad_norm": 0.50390625,
      "learning_rate": 0.000255752427184466,
      "loss": 0.7713,
      "step": 1844
    },
    {
      "epoch": 0.7454168981364577,
      "grad_norm": 0.5,
      "learning_rate": 0.0002557281553398058,
      "loss": 0.7639,
      "step": 1845
    },
    {
      "epoch": 0.7458209181354477,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002557038834951456,
      "loss": 0.7837,
      "step": 1846
    },
    {
      "epoch": 0.7462249381344377,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002556796116504854,
      "loss": 0.8214,
      "step": 1847
    },
    {
      "epoch": 0.7466289581334276,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002556553398058252,
      "loss": 0.8532,
      "step": 1848
    },
    {
      "epoch": 0.7470329781324175,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00025563106796116505,
      "loss": 0.93,
      "step": 1849
    },
    {
      "epoch": 0.7474369981314075,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002556067961165048,
      "loss": 0.8245,
      "step": 1850
    },
    {
      "epoch": 0.7478410181303975,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002555825242718446,
      "loss": 0.9363,
      "step": 1851
    },
    {
      "epoch": 0.7482450381293874,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00025555825242718443,
      "loss": 0.7932,
      "step": 1852
    },
    {
      "epoch": 0.7486490581283773,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00025553398058252426,
      "loss": 0.8714,
      "step": 1853
    },
    {
      "epoch": 0.7490530781273673,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002555097087378641,
      "loss": 0.7787,
      "step": 1854
    },
    {
      "epoch": 0.7494570981263573,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00025548543689320386,
      "loss": 0.8465,
      "step": 1855
    },
    {
      "epoch": 0.7498611181253472,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00025546116504854364,
      "loss": 0.7784,
      "step": 1856
    },
    {
      "epoch": 0.7502651381243371,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00025543689320388347,
      "loss": 0.8079,
      "step": 1857
    },
    {
      "epoch": 0.7506691581233271,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002554126213592233,
      "loss": 1.0206,
      "step": 1858
    },
    {
      "epoch": 0.7510731781223171,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00025538834951456307,
      "loss": 0.8244,
      "step": 1859
    },
    {
      "epoch": 0.751477198121307,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002553640776699029,
      "loss": 0.7088,
      "step": 1860
    },
    {
      "epoch": 0.751881218120297,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002553398058252427,
      "loss": 0.8365,
      "step": 1861
    },
    {
      "epoch": 0.7522852381192869,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002553155339805825,
      "loss": 0.8639,
      "step": 1862
    },
    {
      "epoch": 0.7526892581182768,
      "grad_norm": 0.625,
      "learning_rate": 0.00025529126213592234,
      "loss": 0.8707,
      "step": 1863
    },
    {
      "epoch": 0.7530932781172668,
      "grad_norm": 0.458984375,
      "learning_rate": 0.0002552669902912621,
      "loss": 0.7749,
      "step": 1864
    },
    {
      "epoch": 0.7534972981162568,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00025524271844660194,
      "loss": 0.8179,
      "step": 1865
    },
    {
      "epoch": 0.7539013181152467,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002552184466019417,
      "loss": 0.8884,
      "step": 1866
    },
    {
      "epoch": 0.7543053381142366,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00025519417475728155,
      "loss": 1.0114,
      "step": 1867
    },
    {
      "epoch": 0.7547093581132266,
      "grad_norm": 0.4453125,
      "learning_rate": 0.0002551699029126214,
      "loss": 0.7617,
      "step": 1868
    },
    {
      "epoch": 0.7551133781122166,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00025514563106796115,
      "loss": 0.8995,
      "step": 1869
    },
    {
      "epoch": 0.7555173981112066,
      "grad_norm": 0.5,
      "learning_rate": 0.0002551213592233009,
      "loss": 0.9006,
      "step": 1870
    },
    {
      "epoch": 0.7559214181101964,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00025509708737864075,
      "loss": 0.786,
      "step": 1871
    },
    {
      "epoch": 0.7563254381091864,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00025507281553398053,
      "loss": 0.8471,
      "step": 1872
    },
    {
      "epoch": 0.7567294581081764,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00025504854368932036,
      "loss": 0.7444,
      "step": 1873
    },
    {
      "epoch": 0.7571334781071664,
      "grad_norm": 0.625,
      "learning_rate": 0.0002550242718446602,
      "loss": 0.9541,
      "step": 1874
    },
    {
      "epoch": 0.7575374981061562,
      "grad_norm": 0.53125,
      "learning_rate": 0.00025499999999999996,
      "loss": 0.7789,
      "step": 1875
    },
    {
      "epoch": 0.7579415181051462,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002549757281553398,
      "loss": 0.8566,
      "step": 1876
    },
    {
      "epoch": 0.7583455381041362,
      "grad_norm": 0.5,
      "learning_rate": 0.00025495145631067957,
      "loss": 0.853,
      "step": 1877
    },
    {
      "epoch": 0.758749558103126,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002549271844660194,
      "loss": 0.8337,
      "step": 1878
    },
    {
      "epoch": 0.759153578102116,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00025490291262135923,
      "loss": 0.8475,
      "step": 1879
    },
    {
      "epoch": 0.759557598101106,
      "grad_norm": 0.55078125,
      "learning_rate": 0.000254878640776699,
      "loss": 0.8532,
      "step": 1880
    },
    {
      "epoch": 0.759961618100096,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002548543689320388,
      "loss": 0.8091,
      "step": 1881
    },
    {
      "epoch": 0.7603656380990859,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002548300970873786,
      "loss": 0.7661,
      "step": 1882
    },
    {
      "epoch": 0.7607696580980758,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00025480582524271844,
      "loss": 0.7983,
      "step": 1883
    },
    {
      "epoch": 0.7611736780970658,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002547815533980582,
      "loss": 0.911,
      "step": 1884
    },
    {
      "epoch": 0.7615776980960558,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00025475728155339804,
      "loss": 0.7735,
      "step": 1885
    },
    {
      "epoch": 0.7619817180950457,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002547330097087378,
      "loss": 0.8211,
      "step": 1886
    },
    {
      "epoch": 0.7623857380940356,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00025470873786407765,
      "loss": 0.729,
      "step": 1887
    },
    {
      "epoch": 0.7627897580930256,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002546844660194175,
      "loss": 0.8824,
      "step": 1888
    },
    {
      "epoch": 0.7631937780920156,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00025466019417475725,
      "loss": 0.7683,
      "step": 1889
    },
    {
      "epoch": 0.7635977980910055,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002546359223300971,
      "loss": 0.8902,
      "step": 1890
    },
    {
      "epoch": 0.7640018180899955,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00025461165048543686,
      "loss": 0.8277,
      "step": 1891
    },
    {
      "epoch": 0.7644058380889854,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002545873786407767,
      "loss": 0.8123,
      "step": 1892
    },
    {
      "epoch": 0.7648098580879753,
      "grad_norm": 0.470703125,
      "learning_rate": 0.0002545631067961165,
      "loss": 0.7491,
      "step": 1893
    },
    {
      "epoch": 0.7652138780869653,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002545388349514563,
      "loss": 0.811,
      "step": 1894
    },
    {
      "epoch": 0.7656178980859553,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002545145631067961,
      "loss": 0.856,
      "step": 1895
    },
    {
      "epoch": 0.7660219180849452,
      "grad_norm": 0.484375,
      "learning_rate": 0.0002544902912621359,
      "loss": 0.7683,
      "step": 1896
    },
    {
      "epoch": 0.7664259380839351,
      "grad_norm": 0.4609375,
      "learning_rate": 0.0002544660194174757,
      "loss": 0.6771,
      "step": 1897
    },
    {
      "epoch": 0.7668299580829251,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002544417475728155,
      "loss": 0.8451,
      "step": 1898
    },
    {
      "epoch": 0.7672339780819151,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00025441747572815533,
      "loss": 1.0596,
      "step": 1899
    },
    {
      "epoch": 0.767637998080905,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002543932038834951,
      "loss": 0.8038,
      "step": 1900
    },
    {
      "epoch": 0.7680420180798949,
      "grad_norm": 0.65625,
      "learning_rate": 0.00025436893203883493,
      "loss": 0.835,
      "step": 1901
    },
    {
      "epoch": 0.7684460380788849,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002543446601941747,
      "loss": 0.8118,
      "step": 1902
    },
    {
      "epoch": 0.7688500580778749,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00025432038834951454,
      "loss": 0.8567,
      "step": 1903
    },
    {
      "epoch": 0.7692540780768649,
      "grad_norm": 0.53125,
      "learning_rate": 0.00025429611650485437,
      "loss": 0.7365,
      "step": 1904
    },
    {
      "epoch": 0.7696580980758547,
      "grad_norm": 0.5,
      "learning_rate": 0.00025427184466019414,
      "loss": 0.8271,
      "step": 1905
    },
    {
      "epoch": 0.7700621180748447,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00025424757281553397,
      "loss": 0.7548,
      "step": 1906
    },
    {
      "epoch": 0.7704661380738347,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00025422330097087375,
      "loss": 0.9258,
      "step": 1907
    },
    {
      "epoch": 0.7708701580728246,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002541990291262136,
      "loss": 0.9004,
      "step": 1908
    },
    {
      "epoch": 0.7712741780718145,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002541747572815534,
      "loss": 0.9835,
      "step": 1909
    },
    {
      "epoch": 0.7716781980708045,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002541504854368932,
      "loss": 0.8316,
      "step": 1910
    },
    {
      "epoch": 0.7720822180697945,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00025412621359223296,
      "loss": 0.9146,
      "step": 1911
    },
    {
      "epoch": 0.7724862380687844,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002541019417475728,
      "loss": 0.9122,
      "step": 1912
    },
    {
      "epoch": 0.7728902580677743,
      "grad_norm": 0.703125,
      "learning_rate": 0.0002540776699029126,
      "loss": 0.8863,
      "step": 1913
    },
    {
      "epoch": 0.7732942780667643,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002540533980582524,
      "loss": 0.8385,
      "step": 1914
    },
    {
      "epoch": 0.7736982980657543,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002540291262135922,
      "loss": 0.8125,
      "step": 1915
    },
    {
      "epoch": 0.7741023180647442,
      "grad_norm": 0.62890625,
      "learning_rate": 0.000254004854368932,
      "loss": 0.8991,
      "step": 1916
    },
    {
      "epoch": 0.7745063380637341,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0002539805825242718,
      "loss": 0.7323,
      "step": 1917
    },
    {
      "epoch": 0.7749103580627241,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00025395631067961165,
      "loss": 0.9481,
      "step": 1918
    },
    {
      "epoch": 0.7753143780617141,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00025393203883495143,
      "loss": 0.94,
      "step": 1919
    },
    {
      "epoch": 0.775718398060704,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00025390776699029126,
      "loss": 0.7283,
      "step": 1920
    },
    {
      "epoch": 0.776122418059694,
      "grad_norm": 0.515625,
      "learning_rate": 0.00025388349514563103,
      "loss": 0.8425,
      "step": 1921
    },
    {
      "epoch": 0.7765264380586839,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00025385922330097086,
      "loss": 0.9606,
      "step": 1922
    },
    {
      "epoch": 0.7769304580576738,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00025383495145631064,
      "loss": 0.8344,
      "step": 1923
    },
    {
      "epoch": 0.7773344780566638,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00025381067961165047,
      "loss": 0.8516,
      "step": 1924
    },
    {
      "epoch": 0.7777384980556538,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002537864077669903,
      "loss": 0.8352,
      "step": 1925
    },
    {
      "epoch": 0.7781425180546437,
      "grad_norm": 0.47265625,
      "learning_rate": 0.0002537621359223301,
      "loss": 0.8973,
      "step": 1926
    },
    {
      "epoch": 0.7785465380536336,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025373786407766985,
      "loss": 0.9225,
      "step": 1927
    },
    {
      "epoch": 0.7789505580526236,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002537135922330097,
      "loss": 0.8249,
      "step": 1928
    },
    {
      "epoch": 0.7793545780516136,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002536893203883495,
      "loss": 0.843,
      "step": 1929
    },
    {
      "epoch": 0.7797585980506035,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002536650485436893,
      "loss": 0.9307,
      "step": 1930
    },
    {
      "epoch": 0.7801626180495934,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002536407766990291,
      "loss": 0.7809,
      "step": 1931
    },
    {
      "epoch": 0.7805666380485834,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002536165048543689,
      "loss": 0.8802,
      "step": 1932
    },
    {
      "epoch": 0.7809706580475734,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002535922330097087,
      "loss": 0.7848,
      "step": 1933
    },
    {
      "epoch": 0.7813746780465634,
      "grad_norm": 0.65625,
      "learning_rate": 0.00025356796116504855,
      "loss": 0.9166,
      "step": 1934
    },
    {
      "epoch": 0.7817786980455532,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002535436893203883,
      "loss": 0.7757,
      "step": 1935
    },
    {
      "epoch": 0.7821827180445432,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00025351941747572815,
      "loss": 0.8152,
      "step": 1936
    },
    {
      "epoch": 0.7825867380435332,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002534951456310679,
      "loss": 0.8939,
      "step": 1937
    },
    {
      "epoch": 0.782990758042523,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00025347087378640776,
      "loss": 0.7867,
      "step": 1938
    },
    {
      "epoch": 0.783394778041513,
      "grad_norm": 0.484375,
      "learning_rate": 0.0002534466019417476,
      "loss": 0.7182,
      "step": 1939
    },
    {
      "epoch": 0.783798798040503,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00025342233009708736,
      "loss": 0.8763,
      "step": 1940
    },
    {
      "epoch": 0.784202818039493,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00025339805825242714,
      "loss": 1.0015,
      "step": 1941
    },
    {
      "epoch": 0.7846068380384829,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00025337378640776696,
      "loss": 0.8119,
      "step": 1942
    },
    {
      "epoch": 0.7850108580374728,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002533495145631068,
      "loss": 0.9161,
      "step": 1943
    },
    {
      "epoch": 0.7854148780364628,
      "grad_norm": 0.5625,
      "learning_rate": 0.00025332524271844657,
      "loss": 0.7606,
      "step": 1944
    },
    {
      "epoch": 0.7858188980354528,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002533009708737864,
      "loss": 0.7427,
      "step": 1945
    },
    {
      "epoch": 0.7862229180344427,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002532766990291262,
      "loss": 0.8243,
      "step": 1946
    },
    {
      "epoch": 0.7866269380334326,
      "grad_norm": 0.486328125,
      "learning_rate": 0.000253252427184466,
      "loss": 0.7121,
      "step": 1947
    },
    {
      "epoch": 0.7870309580324226,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002532281553398058,
      "loss": 0.7468,
      "step": 1948
    },
    {
      "epoch": 0.7874349780314126,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002532038834951456,
      "loss": 0.8463,
      "step": 1949
    },
    {
      "epoch": 0.7878389980304025,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00025317961165048544,
      "loss": 0.6875,
      "step": 1950
    },
    {
      "epoch": 0.7882430180293925,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002531553398058252,
      "loss": 0.7961,
      "step": 1951
    },
    {
      "epoch": 0.7886470380283824,
      "grad_norm": 0.5390625,
      "learning_rate": 0.000253131067961165,
      "loss": 0.8459,
      "step": 1952
    },
    {
      "epoch": 0.7890510580273724,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002531067961165048,
      "loss": 0.7585,
      "step": 1953
    },
    {
      "epoch": 0.7894550780263623,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00025308252427184465,
      "loss": 0.9844,
      "step": 1954
    },
    {
      "epoch": 0.7898590980253523,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002530582524271845,
      "loss": 0.8868,
      "step": 1955
    },
    {
      "epoch": 0.7902631180243422,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00025303398058252425,
      "loss": 0.7895,
      "step": 1956
    },
    {
      "epoch": 0.7906671380233321,
      "grad_norm": 0.58984375,
      "learning_rate": 0.000253009708737864,
      "loss": 0.7841,
      "step": 1957
    },
    {
      "epoch": 0.7910711580223221,
      "grad_norm": 0.484375,
      "learning_rate": 0.00025298543689320386,
      "loss": 0.774,
      "step": 1958
    },
    {
      "epoch": 0.7914751780213121,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002529611650485437,
      "loss": 0.7831,
      "step": 1959
    },
    {
      "epoch": 0.791879198020302,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00025293689320388346,
      "loss": 0.8803,
      "step": 1960
    },
    {
      "epoch": 0.7922832180192919,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002529126213592233,
      "loss": 0.8391,
      "step": 1961
    },
    {
      "epoch": 0.7926872380182819,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00025288834951456307,
      "loss": 0.7628,
      "step": 1962
    },
    {
      "epoch": 0.7930912580172719,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002528640776699029,
      "loss": 0.7577,
      "step": 1963
    },
    {
      "epoch": 0.7934952780162619,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002528398058252427,
      "loss": 0.7054,
      "step": 1964
    },
    {
      "epoch": 0.7938992980152517,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002528155339805825,
      "loss": 0.7381,
      "step": 1965
    },
    {
      "epoch": 0.7943033180142417,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00025279126213592233,
      "loss": 0.8606,
      "step": 1966
    },
    {
      "epoch": 0.7947073380132317,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002527669902912621,
      "loss": 0.8658,
      "step": 1967
    },
    {
      "epoch": 0.7951113580122217,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00025274271844660193,
      "loss": 0.7463,
      "step": 1968
    },
    {
      "epoch": 0.7955153780112115,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00025271844660194176,
      "loss": 0.8618,
      "step": 1969
    },
    {
      "epoch": 0.7959193980102015,
      "grad_norm": 0.5,
      "learning_rate": 0.00025269417475728154,
      "loss": 0.8495,
      "step": 1970
    },
    {
      "epoch": 0.7963234180091915,
      "grad_norm": 0.458984375,
      "learning_rate": 0.0002526699029126213,
      "loss": 0.8656,
      "step": 1971
    },
    {
      "epoch": 0.7967274380081814,
      "grad_norm": 0.5625,
      "learning_rate": 0.00025264563106796114,
      "loss": 0.8092,
      "step": 1972
    },
    {
      "epoch": 0.7971314580071713,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002526213592233009,
      "loss": 0.9838,
      "step": 1973
    },
    {
      "epoch": 0.7975354780061613,
      "grad_norm": 0.53125,
      "learning_rate": 0.00025259708737864075,
      "loss": 0.8973,
      "step": 1974
    },
    {
      "epoch": 0.7979394980051513,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002525728155339806,
      "loss": 0.8577,
      "step": 1975
    },
    {
      "epoch": 0.7983435180041412,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00025254854368932035,
      "loss": 0.8721,
      "step": 1976
    },
    {
      "epoch": 0.7987475380031311,
      "grad_norm": 0.4609375,
      "learning_rate": 0.0002525242718446602,
      "loss": 0.7285,
      "step": 1977
    },
    {
      "epoch": 0.7991515580021211,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00025249999999999996,
      "loss": 0.7906,
      "step": 1978
    },
    {
      "epoch": 0.7995555780011111,
      "grad_norm": 0.462890625,
      "learning_rate": 0.0002524757281553398,
      "loss": 0.6765,
      "step": 1979
    },
    {
      "epoch": 0.799959598000101,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0002524514563106796,
      "loss": 0.7873,
      "step": 1980
    },
    {
      "epoch": 0.800363617999091,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002524271844660194,
      "loss": 0.827,
      "step": 1981
    },
    {
      "epoch": 0.8007676379980809,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00025240291262135917,
      "loss": 0.8464,
      "step": 1982
    },
    {
      "epoch": 0.8011716579970709,
      "grad_norm": 0.55078125,
      "learning_rate": 0.000252378640776699,
      "loss": 0.855,
      "step": 1983
    },
    {
      "epoch": 0.8015756779960608,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002523543689320388,
      "loss": 0.8048,
      "step": 1984
    },
    {
      "epoch": 0.8019796979950508,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002523300970873786,
      "loss": 0.8415,
      "step": 1985
    },
    {
      "epoch": 0.8023837179940407,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00025230582524271843,
      "loss": 0.7625,
      "step": 1986
    },
    {
      "epoch": 0.8027877379930306,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002522815533980582,
      "loss": 0.8774,
      "step": 1987
    },
    {
      "epoch": 0.8031917579920206,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00025225728155339804,
      "loss": 0.8135,
      "step": 1988
    },
    {
      "epoch": 0.8035957779910106,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00025223300970873786,
      "loss": 0.7615,
      "step": 1989
    },
    {
      "epoch": 0.8039997979900005,
      "grad_norm": 0.44140625,
      "learning_rate": 0.00025220873786407764,
      "loss": 0.685,
      "step": 1990
    },
    {
      "epoch": 0.8044038179889904,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00025218446601941747,
      "loss": 0.78,
      "step": 1991
    },
    {
      "epoch": 0.8048078379879804,
      "grad_norm": 0.46875,
      "learning_rate": 0.00025216019417475724,
      "loss": 0.7299,
      "step": 1992
    },
    {
      "epoch": 0.8052118579869704,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002521359223300971,
      "loss": 0.8101,
      "step": 1993
    },
    {
      "epoch": 0.8056158779859603,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002521116504854369,
      "loss": 0.8856,
      "step": 1994
    },
    {
      "epoch": 0.8060198979849502,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002520873786407767,
      "loss": 0.7669,
      "step": 1995
    },
    {
      "epoch": 0.8064239179839402,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002520631067961165,
      "loss": 0.841,
      "step": 1996
    },
    {
      "epoch": 0.8068279379829302,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002520388349514563,
      "loss": 0.8257,
      "step": 1997
    },
    {
      "epoch": 0.8072319579819202,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002520145631067961,
      "loss": 0.8575,
      "step": 1998
    },
    {
      "epoch": 0.80763597798091,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002519902912621359,
      "loss": 0.8621,
      "step": 1999
    },
    {
      "epoch": 0.8080399979799,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002519660194174757,
      "loss": 0.7655,
      "step": 2000
    },
    {
      "epoch": 0.80844401797889,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002519417475728155,
      "loss": 0.9522,
      "step": 2001
    },
    {
      "epoch": 0.8088480379778799,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002519174757281553,
      "loss": 0.887,
      "step": 2002
    },
    {
      "epoch": 0.8092520579768698,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002518932038834951,
      "loss": 0.9039,
      "step": 2003
    },
    {
      "epoch": 0.8096560779758598,
      "grad_norm": 0.453125,
      "learning_rate": 0.0002518689320388349,
      "loss": 0.7313,
      "step": 2004
    },
    {
      "epoch": 0.8100600979748498,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00025184466019417476,
      "loss": 0.9392,
      "step": 2005
    },
    {
      "epoch": 0.8104641179738397,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00025182038834951453,
      "loss": 0.8118,
      "step": 2006
    },
    {
      "epoch": 0.8108681379728296,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00025179611650485436,
      "loss": 0.8801,
      "step": 2007
    },
    {
      "epoch": 0.8112721579718196,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00025177184466019414,
      "loss": 0.9204,
      "step": 2008
    },
    {
      "epoch": 0.8116761779708096,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00025174757281553397,
      "loss": 0.7986,
      "step": 2009
    },
    {
      "epoch": 0.8120801979697995,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002517233009708738,
      "loss": 0.9228,
      "step": 2010
    },
    {
      "epoch": 0.8124842179687894,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00025169902912621357,
      "loss": 0.7954,
      "step": 2011
    },
    {
      "epoch": 0.8128882379677794,
      "grad_norm": 0.5,
      "learning_rate": 0.00025167475728155335,
      "loss": 0.8183,
      "step": 2012
    },
    {
      "epoch": 0.8132922579667694,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002516504854368932,
      "loss": 0.871,
      "step": 2013
    },
    {
      "epoch": 0.8136962779657593,
      "grad_norm": 0.6015625,
      "learning_rate": 0.000251626213592233,
      "loss": 0.726,
      "step": 2014
    },
    {
      "epoch": 0.8141002979647493,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002516019417475728,
      "loss": 0.9704,
      "step": 2015
    },
    {
      "epoch": 0.8145043179637392,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0002515776699029126,
      "loss": 0.8682,
      "step": 2016
    },
    {
      "epoch": 0.8149083379627291,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002515533980582524,
      "loss": 0.7542,
      "step": 2017
    },
    {
      "epoch": 0.8153123579617191,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0002515291262135922,
      "loss": 1.011,
      "step": 2018
    },
    {
      "epoch": 0.8157163779607091,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00025150485436893204,
      "loss": 0.832,
      "step": 2019
    },
    {
      "epoch": 0.816120397959699,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002514805825242718,
      "loss": 0.8639,
      "step": 2020
    },
    {
      "epoch": 0.8165244179586889,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00025145631067961165,
      "loss": 0.8033,
      "step": 2021
    },
    {
      "epoch": 0.8169284379576789,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002514320388349514,
      "loss": 0.8414,
      "step": 2022
    },
    {
      "epoch": 0.8173324579566689,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00025140776699029125,
      "loss": 0.8359,
      "step": 2023
    },
    {
      "epoch": 0.8177364779556588,
      "grad_norm": 0.46875,
      "learning_rate": 0.00025138349514563103,
      "loss": 0.7571,
      "step": 2024
    },
    {
      "epoch": 0.8181404979546487,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00025135922330097086,
      "loss": 1.082,
      "step": 2025
    },
    {
      "epoch": 0.8185445179536387,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002513349514563107,
      "loss": 0.8374,
      "step": 2026
    },
    {
      "epoch": 0.8189485379526287,
      "grad_norm": 0.5,
      "learning_rate": 0.00025131067961165046,
      "loss": 0.8659,
      "step": 2027
    },
    {
      "epoch": 0.8193525579516187,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00025128640776699024,
      "loss": 0.8392,
      "step": 2028
    },
    {
      "epoch": 0.8197565779506085,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00025126213592233007,
      "loss": 0.8356,
      "step": 2029
    },
    {
      "epoch": 0.8201605979495985,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002512378640776699,
      "loss": 0.7584,
      "step": 2030
    },
    {
      "epoch": 0.8205646179485885,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00025121359223300967,
      "loss": 0.7475,
      "step": 2031
    },
    {
      "epoch": 0.8209686379475784,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002511893203883495,
      "loss": 0.8234,
      "step": 2032
    },
    {
      "epoch": 0.8213726579465683,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002511650485436893,
      "loss": 0.8767,
      "step": 2033
    },
    {
      "epoch": 0.8217766779455583,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002511407766990291,
      "loss": 0.752,
      "step": 2034
    },
    {
      "epoch": 0.8221806979445483,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00025111650485436893,
      "loss": 0.8553,
      "step": 2035
    },
    {
      "epoch": 0.8225847179435382,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002510922330097087,
      "loss": 0.8091,
      "step": 2036
    },
    {
      "epoch": 0.8229887379425281,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00025106796116504854,
      "loss": 0.8662,
      "step": 2037
    },
    {
      "epoch": 0.8233927579415181,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0002510436893203883,
      "loss": 0.9187,
      "step": 2038
    },
    {
      "epoch": 0.8237967779405081,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00025101941747572814,
      "loss": 0.786,
      "step": 2039
    },
    {
      "epoch": 0.824200797939498,
      "grad_norm": 0.66015625,
      "learning_rate": 0.000250995145631068,
      "loss": 0.8948,
      "step": 2040
    },
    {
      "epoch": 0.8246048179384879,
      "grad_norm": 0.640625,
      "learning_rate": 0.00025097087378640775,
      "loss": 0.7899,
      "step": 2041
    },
    {
      "epoch": 0.8250088379374779,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002509466019417475,
      "loss": 0.8046,
      "step": 2042
    },
    {
      "epoch": 0.8254128579364679,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00025092233009708735,
      "loss": 0.9616,
      "step": 2043
    },
    {
      "epoch": 0.8258168779354578,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002508980582524272,
      "loss": 0.79,
      "step": 2044
    },
    {
      "epoch": 0.8262208979344478,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00025087378640776696,
      "loss": 0.7974,
      "step": 2045
    },
    {
      "epoch": 0.8266249179334377,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002508495145631068,
      "loss": 0.8337,
      "step": 2046
    },
    {
      "epoch": 0.8270289379324277,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00025082524271844656,
      "loss": 0.8338,
      "step": 2047
    },
    {
      "epoch": 0.8274329579314176,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002508009708737864,
      "loss": 0.7865,
      "step": 2048
    },
    {
      "epoch": 0.8278369779304076,
      "grad_norm": 0.5,
      "learning_rate": 0.00025077669902912617,
      "loss": 0.892,
      "step": 2049
    },
    {
      "epoch": 0.8282409979293975,
      "grad_norm": 0.6171875,
      "learning_rate": 0.000250752427184466,
      "loss": 0.9296,
      "step": 2050
    },
    {
      "epoch": 0.8286450179283874,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002507281553398058,
      "loss": 0.7757,
      "step": 2051
    },
    {
      "epoch": 0.8290490379273774,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002507038834951456,
      "loss": 0.7318,
      "step": 2052
    },
    {
      "epoch": 0.8294530579263674,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002506796116504854,
      "loss": 0.8212,
      "step": 2053
    },
    {
      "epoch": 0.8298570779253573,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002506553398058252,
      "loss": 0.9084,
      "step": 2054
    },
    {
      "epoch": 0.8302610979243472,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00025063106796116504,
      "loss": 0.8986,
      "step": 2055
    },
    {
      "epoch": 0.8306651179233372,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00025060679611650487,
      "loss": 0.8831,
      "step": 2056
    },
    {
      "epoch": 0.8310691379223272,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00025058252427184464,
      "loss": 0.8216,
      "step": 2057
    },
    {
      "epoch": 0.8314731579213172,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002505582524271844,
      "loss": 0.8147,
      "step": 2058
    },
    {
      "epoch": 0.831877177920307,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00025053398058252425,
      "loss": 0.7787,
      "step": 2059
    },
    {
      "epoch": 0.832281197919297,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002505097087378641,
      "loss": 0.7728,
      "step": 2060
    },
    {
      "epoch": 0.832685217918287,
      "grad_norm": 0.71875,
      "learning_rate": 0.00025048543689320385,
      "loss": 1.0049,
      "step": 2061
    },
    {
      "epoch": 0.833089237917277,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002504611650485437,
      "loss": 0.746,
      "step": 2062
    },
    {
      "epoch": 0.8334932579162668,
      "grad_norm": 0.578125,
      "learning_rate": 0.00025043689320388345,
      "loss": 0.9051,
      "step": 2063
    },
    {
      "epoch": 0.8338972779152568,
      "grad_norm": 0.455078125,
      "learning_rate": 0.0002504126213592233,
      "loss": 0.7467,
      "step": 2064
    },
    {
      "epoch": 0.8343012979142468,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002503883495145631,
      "loss": 0.8165,
      "step": 2065
    },
    {
      "epoch": 0.8347053179132367,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002503640776699029,
      "loss": 0.9335,
      "step": 2066
    },
    {
      "epoch": 0.8351093379122266,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002503398058252427,
      "loss": 0.9288,
      "step": 2067
    },
    {
      "epoch": 0.8355133579112166,
      "grad_norm": 0.455078125,
      "learning_rate": 0.0002503155339805825,
      "loss": 0.766,
      "step": 2068
    },
    {
      "epoch": 0.8359173779102066,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002502912621359223,
      "loss": 0.7418,
      "step": 2069
    },
    {
      "epoch": 0.8363213979091965,
      "grad_norm": 0.515625,
      "learning_rate": 0.00025026699029126215,
      "loss": 0.7875,
      "step": 2070
    },
    {
      "epoch": 0.8367254179081864,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00025024271844660193,
      "loss": 0.7943,
      "step": 2071
    },
    {
      "epoch": 0.8371294379071764,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002502184466019417,
      "loss": 0.8219,
      "step": 2072
    },
    {
      "epoch": 0.8375334579061664,
      "grad_norm": 0.609375,
      "learning_rate": 0.00025019417475728153,
      "loss": 0.8787,
      "step": 2073
    },
    {
      "epoch": 0.8379374779051563,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002501699029126213,
      "loss": 0.8422,
      "step": 2074
    },
    {
      "epoch": 0.8383414979041462,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00025014563106796114,
      "loss": 0.8906,
      "step": 2075
    },
    {
      "epoch": 0.8387455179031362,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00025012135922330097,
      "loss": 0.9403,
      "step": 2076
    },
    {
      "epoch": 0.8391495379021262,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00025009708737864074,
      "loss": 0.7219,
      "step": 2077
    },
    {
      "epoch": 0.8395535579011161,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00025007281553398057,
      "loss": 0.942,
      "step": 2078
    },
    {
      "epoch": 0.8399575779001061,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00025004854368932035,
      "loss": 0.7658,
      "step": 2079
    },
    {
      "epoch": 0.840361597899096,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002500242718446602,
      "loss": 0.8327,
      "step": 2080
    },
    {
      "epoch": 0.8407656178980859,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00025,
      "loss": 0.929,
      "step": 2081
    },
    {
      "epoch": 0.8411696378970759,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002499757281553398,
      "loss": 0.8686,
      "step": 2082
    },
    {
      "epoch": 0.8415736578960659,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00024995145631067956,
      "loss": 0.8603,
      "step": 2083
    },
    {
      "epoch": 0.8419776778950558,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002499271844660194,
      "loss": 0.7427,
      "step": 2084
    },
    {
      "epoch": 0.8423816978940457,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002499029126213592,
      "loss": 0.8506,
      "step": 2085
    },
    {
      "epoch": 0.8427857178930357,
      "grad_norm": 0.54296875,
      "learning_rate": 0.000249878640776699,
      "loss": 0.8287,
      "step": 2086
    },
    {
      "epoch": 0.8431897378920257,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0002498543689320388,
      "loss": 1.0063,
      "step": 2087
    },
    {
      "epoch": 0.8435937578910156,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002498300970873786,
      "loss": 0.8335,
      "step": 2088
    },
    {
      "epoch": 0.8439977778900055,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002498058252427184,
      "loss": 0.7714,
      "step": 2089
    },
    {
      "epoch": 0.8444017978889955,
      "grad_norm": 0.609375,
      "learning_rate": 0.00024978155339805825,
      "loss": 0.8799,
      "step": 2090
    },
    {
      "epoch": 0.8448058178879855,
      "grad_norm": 0.5,
      "learning_rate": 0.00024975728155339803,
      "loss": 0.7935,
      "step": 2091
    },
    {
      "epoch": 0.8452098378869755,
      "grad_norm": 0.5625,
      "learning_rate": 0.00024973300970873786,
      "loss": 0.854,
      "step": 2092
    },
    {
      "epoch": 0.8456138578859653,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00024970873786407763,
      "loss": 0.8619,
      "step": 2093
    },
    {
      "epoch": 0.8460178778849553,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00024968446601941746,
      "loss": 0.9443,
      "step": 2094
    },
    {
      "epoch": 0.8464218978839453,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002496601941747573,
      "loss": 0.785,
      "step": 2095
    },
    {
      "epoch": 0.8468259178829352,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00024963592233009707,
      "loss": 0.8044,
      "step": 2096
    },
    {
      "epoch": 0.8472299378819251,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002496116504854369,
      "loss": 0.8722,
      "step": 2097
    },
    {
      "epoch": 0.8476339578809151,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00024958737864077667,
      "loss": 0.8358,
      "step": 2098
    },
    {
      "epoch": 0.8480379778799051,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00024956310679611645,
      "loss": 0.8329,
      "step": 2099
    },
    {
      "epoch": 0.848441997878895,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002495388349514563,
      "loss": 0.7529,
      "step": 2100
    },
    {
      "epoch": 0.8488460178778849,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002495145631067961,
      "loss": 0.8119,
      "step": 2101
    },
    {
      "epoch": 0.8492500378768749,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002494902912621359,
      "loss": 0.7948,
      "step": 2102
    },
    {
      "epoch": 0.8496540578758649,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002494660194174757,
      "loss": 0.9167,
      "step": 2103
    },
    {
      "epoch": 0.8500580778748548,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002494417475728155,
      "loss": 0.8221,
      "step": 2104
    },
    {
      "epoch": 0.8504620978738447,
      "grad_norm": 0.439453125,
      "learning_rate": 0.0002494174757281553,
      "loss": 0.7619,
      "step": 2105
    },
    {
      "epoch": 0.8508661178728347,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024939320388349515,
      "loss": 0.7957,
      "step": 2106
    },
    {
      "epoch": 0.8512701378718247,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002493689320388349,
      "loss": 0.8636,
      "step": 2107
    },
    {
      "epoch": 0.8516741578708146,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024934466019417475,
      "loss": 0.7903,
      "step": 2108
    },
    {
      "epoch": 0.8520781778698046,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002493203883495145,
      "loss": 0.8143,
      "step": 2109
    },
    {
      "epoch": 0.8524821978687945,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00024929611650485435,
      "loss": 0.9263,
      "step": 2110
    },
    {
      "epoch": 0.8528862178677844,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002492718446601942,
      "loss": 0.7413,
      "step": 2111
    },
    {
      "epoch": 0.8532902378667744,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024924757281553396,
      "loss": 0.8778,
      "step": 2112
    },
    {
      "epoch": 0.8536942578657644,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00024922330097087373,
      "loss": 0.7576,
      "step": 2113
    },
    {
      "epoch": 0.8540982778647543,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00024919902912621356,
      "loss": 0.7962,
      "step": 2114
    },
    {
      "epoch": 0.8545022978637442,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002491747572815534,
      "loss": 0.8293,
      "step": 2115
    },
    {
      "epoch": 0.8549063178627342,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00024915048543689317,
      "loss": 0.848,
      "step": 2116
    },
    {
      "epoch": 0.8553103378617242,
      "grad_norm": 0.5703125,
      "learning_rate": 0.000249126213592233,
      "loss": 0.8651,
      "step": 2117
    },
    {
      "epoch": 0.8557143578607141,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002491019417475728,
      "loss": 0.8458,
      "step": 2118
    },
    {
      "epoch": 0.856118377859704,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002490776699029126,
      "loss": 0.7742,
      "step": 2119
    },
    {
      "epoch": 0.856522397858694,
      "grad_norm": 0.4375,
      "learning_rate": 0.00024905339805825243,
      "loss": 0.6913,
      "step": 2120
    },
    {
      "epoch": 0.856926417857684,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002490291262135922,
      "loss": 0.8009,
      "step": 2121
    },
    {
      "epoch": 0.857330437856674,
      "grad_norm": 0.5625,
      "learning_rate": 0.00024900485436893204,
      "loss": 0.7903,
      "step": 2122
    },
    {
      "epoch": 0.8577344578556638,
      "grad_norm": 0.46875,
      "learning_rate": 0.0002489805825242718,
      "loss": 0.7584,
      "step": 2123
    },
    {
      "epoch": 0.8581384778546538,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00024895631067961164,
      "loss": 0.9547,
      "step": 2124
    },
    {
      "epoch": 0.8585424978536438,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002489320388349514,
      "loss": 0.9224,
      "step": 2125
    },
    {
      "epoch": 0.8589465178526337,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00024890776699029125,
      "loss": 0.782,
      "step": 2126
    },
    {
      "epoch": 0.8593505378516236,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002488834951456311,
      "loss": 0.7863,
      "step": 2127
    },
    {
      "epoch": 0.8597545578506136,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00024885922330097085,
      "loss": 0.7901,
      "step": 2128
    },
    {
      "epoch": 0.8601585778496036,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002488349514563106,
      "loss": 0.8203,
      "step": 2129
    },
    {
      "epoch": 0.8605625978485935,
      "grad_norm": 0.46875,
      "learning_rate": 0.00024881067961165046,
      "loss": 0.7943,
      "step": 2130
    },
    {
      "epoch": 0.8609666178475834,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002487864077669903,
      "loss": 0.912,
      "step": 2131
    },
    {
      "epoch": 0.8613706378465734,
      "grad_norm": 0.671875,
      "learning_rate": 0.00024876213592233006,
      "loss": 0.8662,
      "step": 2132
    },
    {
      "epoch": 0.8617746578455634,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002487378640776699,
      "loss": 0.8475,
      "step": 2133
    },
    {
      "epoch": 0.8621786778445533,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00024871359223300966,
      "loss": 0.7681,
      "step": 2134
    },
    {
      "epoch": 0.8625826978435432,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002486893203883495,
      "loss": 0.7617,
      "step": 2135
    },
    {
      "epoch": 0.8629867178425332,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002486650485436893,
      "loss": 0.8166,
      "step": 2136
    },
    {
      "epoch": 0.8633907378415232,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002486407766990291,
      "loss": 0.7719,
      "step": 2137
    },
    {
      "epoch": 0.8637947578405131,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00024861650485436893,
      "loss": 0.8793,
      "step": 2138
    },
    {
      "epoch": 0.864198777839503,
      "grad_norm": 0.48046875,
      "learning_rate": 0.0002485922330097087,
      "loss": 0.7441,
      "step": 2139
    },
    {
      "epoch": 0.864602797838493,
      "grad_norm": 0.53125,
      "learning_rate": 0.00024856796116504853,
      "loss": 0.6972,
      "step": 2140
    },
    {
      "epoch": 0.865006817837483,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00024854368932038836,
      "loss": 0.8296,
      "step": 2141
    },
    {
      "epoch": 0.8654108378364729,
      "grad_norm": 0.46875,
      "learning_rate": 0.00024851941747572814,
      "loss": 0.8329,
      "step": 2142
    },
    {
      "epoch": 0.8658148578354629,
      "grad_norm": 0.470703125,
      "learning_rate": 0.0002484951456310679,
      "loss": 0.805,
      "step": 2143
    },
    {
      "epoch": 0.8662188778344528,
      "grad_norm": 0.484375,
      "learning_rate": 0.00024847087378640774,
      "loss": 0.7829,
      "step": 2144
    },
    {
      "epoch": 0.8666228978334427,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00024844660194174757,
      "loss": 0.8937,
      "step": 2145
    },
    {
      "epoch": 0.8670269178324327,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00024842233009708735,
      "loss": 0.7418,
      "step": 2146
    },
    {
      "epoch": 0.8674309378314227,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002483980582524272,
      "loss": 0.8781,
      "step": 2147
    },
    {
      "epoch": 0.8678349578304126,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00024837378640776695,
      "loss": 0.7678,
      "step": 2148
    },
    {
      "epoch": 0.8682389778294025,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002483495145631068,
      "loss": 0.7625,
      "step": 2149
    },
    {
      "epoch": 0.8686429978283925,
      "grad_norm": 0.484375,
      "learning_rate": 0.00024832524271844656,
      "loss": 0.7013,
      "step": 2150
    },
    {
      "epoch": 0.8690470178273825,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002483009708737864,
      "loss": 0.7965,
      "step": 2151
    },
    {
      "epoch": 0.8694510378263725,
      "grad_norm": 0.478515625,
      "learning_rate": 0.0002482766990291262,
      "loss": 0.7895,
      "step": 2152
    },
    {
      "epoch": 0.8698550578253623,
      "grad_norm": 0.55859375,
      "learning_rate": 0.000248252427184466,
      "loss": 0.8216,
      "step": 2153
    },
    {
      "epoch": 0.8702590778243523,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00024822815533980577,
      "loss": 0.735,
      "step": 2154
    },
    {
      "epoch": 0.8706630978233423,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002482038834951456,
      "loss": 0.7698,
      "step": 2155
    },
    {
      "epoch": 0.8710671178223323,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002481796116504854,
      "loss": 0.9599,
      "step": 2156
    },
    {
      "epoch": 0.8714711378213221,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00024815533980582525,
      "loss": 0.8189,
      "step": 2157
    },
    {
      "epoch": 0.8718751578203121,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00024813106796116503,
      "loss": 0.8148,
      "step": 2158
    },
    {
      "epoch": 0.8722791778193021,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002481067961165048,
      "loss": 0.8466,
      "step": 2159
    },
    {
      "epoch": 0.872683197818292,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00024808252427184463,
      "loss": 0.9533,
      "step": 2160
    },
    {
      "epoch": 0.8730872178172819,
      "grad_norm": 0.53125,
      "learning_rate": 0.00024805825242718446,
      "loss": 0.6904,
      "step": 2161
    },
    {
      "epoch": 0.8734912378162719,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00024803398058252424,
      "loss": 0.6946,
      "step": 2162
    },
    {
      "epoch": 0.8738952578152619,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00024800970873786407,
      "loss": 0.8714,
      "step": 2163
    },
    {
      "epoch": 0.8742992778142518,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00024798543689320384,
      "loss": 0.7367,
      "step": 2164
    },
    {
      "epoch": 0.8747032978132417,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00024796116504854367,
      "loss": 0.8131,
      "step": 2165
    },
    {
      "epoch": 0.8751073178122317,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002479368932038835,
      "loss": 0.7071,
      "step": 2166
    },
    {
      "epoch": 0.8755113378112217,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002479126213592233,
      "loss": 0.8652,
      "step": 2167
    },
    {
      "epoch": 0.8759153578102116,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002478883495145631,
      "loss": 0.8671,
      "step": 2168
    },
    {
      "epoch": 0.8763193778092015,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002478640776699029,
      "loss": 0.8046,
      "step": 2169
    },
    {
      "epoch": 0.8767233978081915,
      "grad_norm": 0.5,
      "learning_rate": 0.0002478398058252427,
      "loss": 0.7597,
      "step": 2170
    },
    {
      "epoch": 0.8771274178071815,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00024781553398058254,
      "loss": 0.7525,
      "step": 2171
    },
    {
      "epoch": 0.8775314378061714,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002477912621359223,
      "loss": 0.7309,
      "step": 2172
    },
    {
      "epoch": 0.8779354578051614,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0002477669902912621,
      "loss": 0.9239,
      "step": 2173
    },
    {
      "epoch": 0.8783394778041513,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002477427184466019,
      "loss": 0.897,
      "step": 2174
    },
    {
      "epoch": 0.8787434978031412,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002477184466019417,
      "loss": 0.816,
      "step": 2175
    },
    {
      "epoch": 0.8791475178021312,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002476941747572815,
      "loss": 0.9232,
      "step": 2176
    },
    {
      "epoch": 0.8795515378011212,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00024766990291262136,
      "loss": 0.8328,
      "step": 2177
    },
    {
      "epoch": 0.8799555578001111,
      "grad_norm": 0.453125,
      "learning_rate": 0.00024764563106796113,
      "loss": 0.7356,
      "step": 2178
    },
    {
      "epoch": 0.880359577799101,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00024762135922330096,
      "loss": 0.9307,
      "step": 2179
    },
    {
      "epoch": 0.880763597798091,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00024759708737864074,
      "loss": 0.7461,
      "step": 2180
    },
    {
      "epoch": 0.881167617797081,
      "grad_norm": 0.43359375,
      "learning_rate": 0.00024757281553398056,
      "loss": 0.8155,
      "step": 2181
    },
    {
      "epoch": 0.881571637796071,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002475485436893204,
      "loss": 0.9277,
      "step": 2182
    },
    {
      "epoch": 0.8819756577950608,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00024752427184466017,
      "loss": 0.8649,
      "step": 2183
    },
    {
      "epoch": 0.8823796777940508,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00024749999999999994,
      "loss": 0.8771,
      "step": 2184
    },
    {
      "epoch": 0.8827836977930408,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002474757281553398,
      "loss": 0.774,
      "step": 2185
    },
    {
      "epoch": 0.8831877177920308,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002474514563106796,
      "loss": 0.852,
      "step": 2186
    },
    {
      "epoch": 0.8835917377910206,
      "grad_norm": 0.46875,
      "learning_rate": 0.0002474271844660194,
      "loss": 0.7961,
      "step": 2187
    },
    {
      "epoch": 0.8839957577900106,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002474029126213592,
      "loss": 0.7441,
      "step": 2188
    },
    {
      "epoch": 0.8843997777890006,
      "grad_norm": 0.5078125,
      "learning_rate": 0.000247378640776699,
      "loss": 0.8155,
      "step": 2189
    },
    {
      "epoch": 0.8848037977879905,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002473543689320388,
      "loss": 0.8531,
      "step": 2190
    },
    {
      "epoch": 0.8852078177869804,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00024733009708737864,
      "loss": 0.7841,
      "step": 2191
    },
    {
      "epoch": 0.8856118377859704,
      "grad_norm": 0.46484375,
      "learning_rate": 0.0002473058252427184,
      "loss": 0.7612,
      "step": 2192
    },
    {
      "epoch": 0.8860158577849604,
      "grad_norm": 0.671875,
      "learning_rate": 0.00024728155339805825,
      "loss": 0.8863,
      "step": 2193
    },
    {
      "epoch": 0.8864198777839503,
      "grad_norm": 0.486328125,
      "learning_rate": 0.000247257281553398,
      "loss": 0.8257,
      "step": 2194
    },
    {
      "epoch": 0.8868238977829402,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00024723300970873785,
      "loss": 0.8744,
      "step": 2195
    },
    {
      "epoch": 0.8872279177819302,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0002472087378640777,
      "loss": 0.7236,
      "step": 2196
    },
    {
      "epoch": 0.8876319377809202,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00024718446601941746,
      "loss": 0.8046,
      "step": 2197
    },
    {
      "epoch": 0.8880359577799101,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002471601941747573,
      "loss": 0.8376,
      "step": 2198
    },
    {
      "epoch": 0.8884399777789,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00024713592233009706,
      "loss": 0.8666,
      "step": 2199
    },
    {
      "epoch": 0.88884399777789,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00024711165048543684,
      "loss": 0.8913,
      "step": 2200
    },
    {
      "epoch": 0.88924801777688,
      "grad_norm": 0.5625,
      "learning_rate": 0.00024708737864077667,
      "loss": 0.8095,
      "step": 2201
    },
    {
      "epoch": 0.8896520377758699,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002470631067961165,
      "loss": 0.8852,
      "step": 2202
    },
    {
      "epoch": 0.8900560577748599,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00024703883495145627,
      "loss": 0.8554,
      "step": 2203
    },
    {
      "epoch": 0.8904600777738498,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002470145631067961,
      "loss": 0.6999,
      "step": 2204
    },
    {
      "epoch": 0.8908640977728397,
      "grad_norm": 0.5,
      "learning_rate": 0.0002469902912621359,
      "loss": 0.7925,
      "step": 2205
    },
    {
      "epoch": 0.8912681177718297,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002469660194174757,
      "loss": 0.8087,
      "step": 2206
    },
    {
      "epoch": 0.8916721377708197,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00024694174757281553,
      "loss": 0.8497,
      "step": 2207
    },
    {
      "epoch": 0.8920761577698096,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0002469174757281553,
      "loss": 0.9606,
      "step": 2208
    },
    {
      "epoch": 0.8924801777687995,
      "grad_norm": 0.453125,
      "learning_rate": 0.00024689320388349514,
      "loss": 0.7019,
      "step": 2209
    },
    {
      "epoch": 0.8928841977677895,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002468689320388349,
      "loss": 0.7827,
      "step": 2210
    },
    {
      "epoch": 0.8932882177667795,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00024684466019417474,
      "loss": 0.7655,
      "step": 2211
    },
    {
      "epoch": 0.8936922377657694,
      "grad_norm": 0.65625,
      "learning_rate": 0.00024682038834951457,
      "loss": 1.0427,
      "step": 2212
    },
    {
      "epoch": 0.8940962577647593,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00024679611650485435,
      "loss": 0.8762,
      "step": 2213
    },
    {
      "epoch": 0.8945002777637493,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002467718446601941,
      "loss": 0.8468,
      "step": 2214
    },
    {
      "epoch": 0.8949042977627393,
      "grad_norm": 0.5625,
      "learning_rate": 0.00024674757281553395,
      "loss": 0.906,
      "step": 2215
    },
    {
      "epoch": 0.8953083177617293,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002467233009708738,
      "loss": 0.8247,
      "step": 2216
    },
    {
      "epoch": 0.8957123377607191,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00024669902912621356,
      "loss": 0.7538,
      "step": 2217
    },
    {
      "epoch": 0.8961163577597091,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002466747572815534,
      "loss": 0.8855,
      "step": 2218
    },
    {
      "epoch": 0.8965203777586991,
      "grad_norm": 0.578125,
      "learning_rate": 0.00024665048543689316,
      "loss": 0.8608,
      "step": 2219
    },
    {
      "epoch": 0.896924397757689,
      "grad_norm": 0.578125,
      "learning_rate": 0.000246626213592233,
      "loss": 0.8328,
      "step": 2220
    },
    {
      "epoch": 0.8973284177566789,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002466019417475728,
      "loss": 0.8033,
      "step": 2221
    },
    {
      "epoch": 0.8977324377556689,
      "grad_norm": 0.447265625,
      "learning_rate": 0.0002465776699029126,
      "loss": 0.8049,
      "step": 2222
    },
    {
      "epoch": 0.8981364577546589,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002465533980582524,
      "loss": 0.8108,
      "step": 2223
    },
    {
      "epoch": 0.8985404777536488,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002465291262135922,
      "loss": 0.8753,
      "step": 2224
    },
    {
      "epoch": 0.8989444977526387,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00024650485436893203,
      "loss": 0.8558,
      "step": 2225
    },
    {
      "epoch": 0.8993485177516287,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002464805825242718,
      "loss": 0.8319,
      "step": 2226
    },
    {
      "epoch": 0.8997525377506187,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00024645631067961163,
      "loss": 0.7316,
      "step": 2227
    },
    {
      "epoch": 0.9001565577496086,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00024643203883495146,
      "loss": 0.8059,
      "step": 2228
    },
    {
      "epoch": 0.9005605777485985,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00024640776699029124,
      "loss": 0.8344,
      "step": 2229
    },
    {
      "epoch": 0.9009645977475885,
      "grad_norm": 0.447265625,
      "learning_rate": 0.000246383495145631,
      "loss": 0.8563,
      "step": 2230
    },
    {
      "epoch": 0.9013686177465785,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00024635922330097084,
      "loss": 0.8364,
      "step": 2231
    },
    {
      "epoch": 0.9017726377455684,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002463349514563107,
      "loss": 0.7385,
      "step": 2232
    },
    {
      "epoch": 0.9021766577445584,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00024631067961165045,
      "loss": 0.7999,
      "step": 2233
    },
    {
      "epoch": 0.9025806777435483,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002462864077669903,
      "loss": 0.8091,
      "step": 2234
    },
    {
      "epoch": 0.9029846977425382,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00024626213592233005,
      "loss": 0.8858,
      "step": 2235
    },
    {
      "epoch": 0.9033887177415282,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002462378640776699,
      "loss": 0.8192,
      "step": 2236
    },
    {
      "epoch": 0.9037927377405182,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002462135922330097,
      "loss": 0.8104,
      "step": 2237
    },
    {
      "epoch": 0.9041967577395081,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002461893203883495,
      "loss": 0.9714,
      "step": 2238
    },
    {
      "epoch": 0.904600777738498,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0002461650485436893,
      "loss": 0.7774,
      "step": 2239
    },
    {
      "epoch": 0.905004797737488,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002461407766990291,
      "loss": 0.7746,
      "step": 2240
    },
    {
      "epoch": 0.905408817736478,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002461165048543689,
      "loss": 0.7923,
      "step": 2241
    },
    {
      "epoch": 0.905812837735468,
      "grad_norm": 0.59375,
      "learning_rate": 0.00024609223300970875,
      "loss": 0.9247,
      "step": 2242
    },
    {
      "epoch": 0.9062168577344578,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002460679611650485,
      "loss": 0.8243,
      "step": 2243
    },
    {
      "epoch": 0.9066208777334478,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002460436893203883,
      "loss": 0.9159,
      "step": 2244
    },
    {
      "epoch": 0.9070248977324378,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00024601941747572813,
      "loss": 0.7041,
      "step": 2245
    },
    {
      "epoch": 0.9074289177314278,
      "grad_norm": 0.59375,
      "learning_rate": 0.00024599514563106796,
      "loss": 0.8102,
      "step": 2246
    },
    {
      "epoch": 0.9078329377304176,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00024597087378640774,
      "loss": 0.7113,
      "step": 2247
    },
    {
      "epoch": 0.9082369577294076,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00024594660194174757,
      "loss": 0.8125,
      "step": 2248
    },
    {
      "epoch": 0.9086409777283976,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00024592233009708734,
      "loss": 0.8871,
      "step": 2249
    },
    {
      "epoch": 0.9090449977273876,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00024589805825242717,
      "loss": 0.9702,
      "step": 2250
    },
    {
      "epoch": 0.9094490177263774,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024587378640776695,
      "loss": 0.8018,
      "step": 2251
    },
    {
      "epoch": 0.9098530377253674,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002458495145631068,
      "loss": 0.9382,
      "step": 2252
    },
    {
      "epoch": 0.9102570577243574,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002458252427184466,
      "loss": 0.8248,
      "step": 2253
    },
    {
      "epoch": 0.9106610777233473,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002458009708737864,
      "loss": 0.808,
      "step": 2254
    },
    {
      "epoch": 0.9110650977223372,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00024577669902912615,
      "loss": 0.7667,
      "step": 2255
    },
    {
      "epoch": 0.9114691177213272,
      "grad_norm": 0.474609375,
      "learning_rate": 0.000245752427184466,
      "loss": 0.784,
      "step": 2256
    },
    {
      "epoch": 0.9118731377203172,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002457281553398058,
      "loss": 0.87,
      "step": 2257
    },
    {
      "epoch": 0.9122771577193071,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00024570388349514564,
      "loss": 0.8249,
      "step": 2258
    },
    {
      "epoch": 0.912681177718297,
      "grad_norm": 0.48046875,
      "learning_rate": 0.0002456796116504854,
      "loss": 0.8377,
      "step": 2259
    },
    {
      "epoch": 0.913085197717287,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002456553398058252,
      "loss": 0.8055,
      "step": 2260
    },
    {
      "epoch": 0.913489217716277,
      "grad_norm": 0.5,
      "learning_rate": 0.000245631067961165,
      "loss": 0.8963,
      "step": 2261
    },
    {
      "epoch": 0.9138932377152669,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00024560679611650485,
      "loss": 0.753,
      "step": 2262
    },
    {
      "epoch": 0.9142972577142569,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00024558252427184463,
      "loss": 0.7734,
      "step": 2263
    },
    {
      "epoch": 0.9147012777132468,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00024555825242718446,
      "loss": 0.7899,
      "step": 2264
    },
    {
      "epoch": 0.9151052977122368,
      "grad_norm": 0.515625,
      "learning_rate": 0.00024553398058252423,
      "loss": 0.7598,
      "step": 2265
    },
    {
      "epoch": 0.9155093177112267,
      "grad_norm": 0.5625,
      "learning_rate": 0.00024550970873786406,
      "loss": 0.8358,
      "step": 2266
    },
    {
      "epoch": 0.9159133377102167,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002454854368932039,
      "loss": 0.8817,
      "step": 2267
    },
    {
      "epoch": 0.9163173577092066,
      "grad_norm": 0.515625,
      "learning_rate": 0.00024546116504854367,
      "loss": 0.8672,
      "step": 2268
    },
    {
      "epoch": 0.9167213777081965,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002454368932038835,
      "loss": 0.87,
      "step": 2269
    },
    {
      "epoch": 0.9171253977071865,
      "grad_norm": 0.5,
      "learning_rate": 0.00024541262135922327,
      "loss": 0.7389,
      "step": 2270
    },
    {
      "epoch": 0.9175294177061765,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002453883495145631,
      "loss": 0.7761,
      "step": 2271
    },
    {
      "epoch": 0.9179334377051664,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00024536407766990293,
      "loss": 0.8494,
      "step": 2272
    },
    {
      "epoch": 0.9183374577041563,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002453398058252427,
      "loss": 0.8415,
      "step": 2273
    },
    {
      "epoch": 0.9187414777031463,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002453155339805825,
      "loss": 0.8298,
      "step": 2274
    },
    {
      "epoch": 0.9191454977021363,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002452912621359223,
      "loss": 0.8386,
      "step": 2275
    },
    {
      "epoch": 0.9195495177011263,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002452669902912621,
      "loss": 0.952,
      "step": 2276
    },
    {
      "epoch": 0.9199535377001161,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002452427184466019,
      "loss": 0.8976,
      "step": 2277
    },
    {
      "epoch": 0.9203575576991061,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00024521844660194174,
      "loss": 0.8303,
      "step": 2278
    },
    {
      "epoch": 0.9207615776980961,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002451941747572815,
      "loss": 0.8572,
      "step": 2279
    },
    {
      "epoch": 0.9211655976970861,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00024516990291262135,
      "loss": 0.7421,
      "step": 2280
    },
    {
      "epoch": 0.9215696176960759,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002451456310679611,
      "loss": 0.8683,
      "step": 2281
    },
    {
      "epoch": 0.9219736376950659,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00024512135922330095,
      "loss": 0.7718,
      "step": 2282
    },
    {
      "epoch": 0.9223776576940559,
      "grad_norm": 0.435546875,
      "learning_rate": 0.0002450970873786408,
      "loss": 0.7586,
      "step": 2283
    },
    {
      "epoch": 0.9227816776930458,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00024507281553398056,
      "loss": 0.9129,
      "step": 2284
    },
    {
      "epoch": 0.9231856976920357,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00024504854368932033,
      "loss": 0.7577,
      "step": 2285
    },
    {
      "epoch": 0.9235897176910257,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00024502427184466016,
      "loss": 0.7433,
      "step": 2286
    },
    {
      "epoch": 0.9239937376900157,
      "grad_norm": 0.55078125,
      "learning_rate": 0.000245,
      "loss": 0.8101,
      "step": 2287
    },
    {
      "epoch": 0.9243977576890056,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00024497572815533977,
      "loss": 0.8149,
      "step": 2288
    },
    {
      "epoch": 0.9248017776879955,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002449514563106796,
      "loss": 0.7963,
      "step": 2289
    },
    {
      "epoch": 0.9252057976869855,
      "grad_norm": 0.53125,
      "learning_rate": 0.00024492718446601937,
      "loss": 0.789,
      "step": 2290
    },
    {
      "epoch": 0.9256098176859755,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002449029126213592,
      "loss": 0.8137,
      "step": 2291
    },
    {
      "epoch": 0.9260138376849654,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00024487864077669903,
      "loss": 0.8352,
      "step": 2292
    },
    {
      "epoch": 0.9264178576839553,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002448543689320388,
      "loss": 0.8374,
      "step": 2293
    },
    {
      "epoch": 0.9268218776829453,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00024483009708737864,
      "loss": 0.7237,
      "step": 2294
    },
    {
      "epoch": 0.9272258976819353,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002448058252427184,
      "loss": 0.7325,
      "step": 2295
    },
    {
      "epoch": 0.9276299176809252,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00024478155339805824,
      "loss": 0.8198,
      "step": 2296
    },
    {
      "epoch": 0.9280339376799152,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00024475728155339807,
      "loss": 0.726,
      "step": 2297
    },
    {
      "epoch": 0.9284379576789051,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00024473300970873784,
      "loss": 0.8109,
      "step": 2298
    },
    {
      "epoch": 0.928841977677895,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002447087378640777,
      "loss": 0.6926,
      "step": 2299
    },
    {
      "epoch": 0.929245997676885,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00024468446601941745,
      "loss": 0.8012,
      "step": 2300
    },
    {
      "epoch": 0.929650017675875,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002446601941747572,
      "loss": 0.8122,
      "step": 2301
    },
    {
      "epoch": 0.9300540376748649,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00024463592233009705,
      "loss": 0.7443,
      "step": 2302
    },
    {
      "epoch": 0.9304580576738548,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002446116504854369,
      "loss": 0.8479,
      "step": 2303
    },
    {
      "epoch": 0.9308620776728448,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00024458737864077666,
      "loss": 0.7338,
      "step": 2304
    },
    {
      "epoch": 0.9312660976718348,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002445631067961165,
      "loss": 0.9079,
      "step": 2305
    },
    {
      "epoch": 0.9316701176708247,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00024453883495145626,
      "loss": 0.8914,
      "step": 2306
    },
    {
      "epoch": 0.9320741376698146,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002445145631067961,
      "loss": 0.7793,
      "step": 2307
    },
    {
      "epoch": 0.9324781576688046,
      "grad_norm": 0.470703125,
      "learning_rate": 0.0002444902912621359,
      "loss": 0.8334,
      "step": 2308
    },
    {
      "epoch": 0.9328821776677946,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002444660194174757,
      "loss": 0.7742,
      "step": 2309
    },
    {
      "epoch": 0.9332861976667846,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00024444174757281553,
      "loss": 0.8011,
      "step": 2310
    },
    {
      "epoch": 0.9336902176657744,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002444174757281553,
      "loss": 0.8831,
      "step": 2311
    },
    {
      "epoch": 0.9340942376647644,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00024439320388349513,
      "loss": 0.7364,
      "step": 2312
    },
    {
      "epoch": 0.9344982576637544,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00024436893203883496,
      "loss": 0.8454,
      "step": 2313
    },
    {
      "epoch": 0.9349022776627443,
      "grad_norm": 0.5,
      "learning_rate": 0.00024434466019417474,
      "loss": 0.8867,
      "step": 2314
    },
    {
      "epoch": 0.9353062976617342,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002443203883495145,
      "loss": 0.743,
      "step": 2315
    },
    {
      "epoch": 0.9357103176607242,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024429611650485434,
      "loss": 0.9102,
      "step": 2316
    },
    {
      "epoch": 0.9361143376597142,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00024427184466019417,
      "loss": 0.8955,
      "step": 2317
    },
    {
      "epoch": 0.9365183576587041,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00024424757281553395,
      "loss": 0.8683,
      "step": 2318
    },
    {
      "epoch": 0.936922377657694,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002442233009708738,
      "loss": 0.8126,
      "step": 2319
    },
    {
      "epoch": 0.937326397656684,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00024419902912621355,
      "loss": 0.8359,
      "step": 2320
    },
    {
      "epoch": 0.937730417655674,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002441747572815534,
      "loss": 0.9271,
      "step": 2321
    },
    {
      "epoch": 0.9381344376546639,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0002441504854368932,
      "loss": 0.7565,
      "step": 2322
    },
    {
      "epoch": 0.9385384576536538,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00024412621359223298,
      "loss": 0.7957,
      "step": 2323
    },
    {
      "epoch": 0.9389424776526438,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00024410194174757281,
      "loss": 0.6883,
      "step": 2324
    },
    {
      "epoch": 0.9393464976516338,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002440776699029126,
      "loss": 0.7711,
      "step": 2325
    },
    {
      "epoch": 0.9397505176506237,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002440533980582524,
      "loss": 0.7066,
      "step": 2326
    },
    {
      "epoch": 0.9401545376496137,
      "grad_norm": 0.625,
      "learning_rate": 0.00024402912621359222,
      "loss": 0.8653,
      "step": 2327
    },
    {
      "epoch": 0.9405585576486036,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00024400485436893202,
      "loss": 0.8369,
      "step": 2328
    },
    {
      "epoch": 0.9409625776475935,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00024398058252427183,
      "loss": 0.6516,
      "step": 2329
    },
    {
      "epoch": 0.9413665976465835,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00024395631067961163,
      "loss": 0.8957,
      "step": 2330
    },
    {
      "epoch": 0.9417706176455735,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00024393203883495143,
      "loss": 0.8387,
      "step": 2331
    },
    {
      "epoch": 0.9421746376445634,
      "grad_norm": 0.53125,
      "learning_rate": 0.00024390776699029126,
      "loss": 0.7192,
      "step": 2332
    },
    {
      "epoch": 0.9425786576435533,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00024388349514563104,
      "loss": 0.754,
      "step": 2333
    },
    {
      "epoch": 0.9429826776425433,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00024385922330097084,
      "loss": 0.647,
      "step": 2334
    },
    {
      "epoch": 0.9433866976415333,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00024383495145631067,
      "loss": 0.8795,
      "step": 2335
    },
    {
      "epoch": 0.9437907176405232,
      "grad_norm": 0.44921875,
      "learning_rate": 0.00024381067961165047,
      "loss": 0.7682,
      "step": 2336
    },
    {
      "epoch": 0.9441947376395131,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00024378640776699024,
      "loss": 0.8812,
      "step": 2337
    },
    {
      "epoch": 0.9445987576385031,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00024376213592233007,
      "loss": 0.7165,
      "step": 2338
    },
    {
      "epoch": 0.9450027776374931,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00024373786407766988,
      "loss": 0.8452,
      "step": 2339
    },
    {
      "epoch": 0.945406797636483,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002437135922330097,
      "loss": 0.8352,
      "step": 2340
    },
    {
      "epoch": 0.9458108176354729,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002436893203883495,
      "loss": 0.7456,
      "step": 2341
    },
    {
      "epoch": 0.9462148376344629,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00024366504854368928,
      "loss": 0.7641,
      "step": 2342
    },
    {
      "epoch": 0.9466188576334529,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002436407766990291,
      "loss": 0.7939,
      "step": 2343
    },
    {
      "epoch": 0.9470228776324429,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00024361650485436892,
      "loss": 0.824,
      "step": 2344
    },
    {
      "epoch": 0.9474268976314327,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002435922330097087,
      "loss": 0.7981,
      "step": 2345
    },
    {
      "epoch": 0.9478309176304227,
      "grad_norm": 0.77734375,
      "learning_rate": 0.00024356796116504852,
      "loss": 0.9573,
      "step": 2346
    },
    {
      "epoch": 0.9482349376294127,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024354368932038832,
      "loss": 0.7942,
      "step": 2347
    },
    {
      "epoch": 0.9486389576284026,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00024351941747572812,
      "loss": 0.8108,
      "step": 2348
    },
    {
      "epoch": 0.9490429776273925,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024349514563106795,
      "loss": 0.7443,
      "step": 2349
    },
    {
      "epoch": 0.9494469976263825,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00024347087378640773,
      "loss": 0.9047,
      "step": 2350
    },
    {
      "epoch": 0.9498510176253725,
      "grad_norm": 0.515625,
      "learning_rate": 0.00024344660194174756,
      "loss": 0.8379,
      "step": 2351
    },
    {
      "epoch": 0.9502550376243624,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024342233009708736,
      "loss": 0.8623,
      "step": 2352
    },
    {
      "epoch": 0.9506590576233523,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00024339805825242716,
      "loss": 0.8022,
      "step": 2353
    },
    {
      "epoch": 0.9510630776223423,
      "grad_norm": 0.46875,
      "learning_rate": 0.000243373786407767,
      "loss": 0.7087,
      "step": 2354
    },
    {
      "epoch": 0.9514670976213323,
      "grad_norm": 0.59375,
      "learning_rate": 0.00024334951456310677,
      "loss": 0.9102,
      "step": 2355
    },
    {
      "epoch": 0.9518711176203222,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00024332524271844657,
      "loss": 0.8093,
      "step": 2356
    },
    {
      "epoch": 0.9522751376193122,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002433009708737864,
      "loss": 0.9117,
      "step": 2357
    },
    {
      "epoch": 0.9526791576183021,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00024327669902912618,
      "loss": 0.9443,
      "step": 2358
    },
    {
      "epoch": 0.9530831776172921,
      "grad_norm": 0.5625,
      "learning_rate": 0.000243252427184466,
      "loss": 0.7972,
      "step": 2359
    },
    {
      "epoch": 0.953487197616282,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002432281553398058,
      "loss": 0.83,
      "step": 2360
    },
    {
      "epoch": 0.953891217615272,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002432038834951456,
      "loss": 0.6922,
      "step": 2361
    },
    {
      "epoch": 0.9542952376142619,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00024317961165048544,
      "loss": 0.7308,
      "step": 2362
    },
    {
      "epoch": 0.9546992576132518,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00024315533980582521,
      "loss": 0.8073,
      "step": 2363
    },
    {
      "epoch": 0.9551032776122418,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00024313106796116502,
      "loss": 0.8674,
      "step": 2364
    },
    {
      "epoch": 0.9555072976112318,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00024310679611650485,
      "loss": 0.7358,
      "step": 2365
    },
    {
      "epoch": 0.9559113176102217,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00024308252427184465,
      "loss": 0.7659,
      "step": 2366
    },
    {
      "epoch": 0.9563153376092116,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00024305825242718442,
      "loss": 0.9964,
      "step": 2367
    },
    {
      "epoch": 0.9567193576082016,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00024303398058252425,
      "loss": 0.8175,
      "step": 2368
    },
    {
      "epoch": 0.9571233776071916,
      "grad_norm": 0.7421875,
      "learning_rate": 0.00024300970873786406,
      "loss": 1.0361,
      "step": 2369
    },
    {
      "epoch": 0.9575273976061816,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00024298543689320388,
      "loss": 0.8798,
      "step": 2370
    },
    {
      "epoch": 0.9579314176051714,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024296116504854366,
      "loss": 0.8493,
      "step": 2371
    },
    {
      "epoch": 0.9583354376041614,
      "grad_norm": 0.5625,
      "learning_rate": 0.00024293689320388346,
      "loss": 0.9715,
      "step": 2372
    },
    {
      "epoch": 0.9587394576031514,
      "grad_norm": 0.5,
      "learning_rate": 0.0002429126213592233,
      "loss": 0.887,
      "step": 2373
    },
    {
      "epoch": 0.9591434776021414,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002428883495145631,
      "loss": 0.7447,
      "step": 2374
    },
    {
      "epoch": 0.9595474976011312,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00024286407766990287,
      "loss": 0.7537,
      "step": 2375
    },
    {
      "epoch": 0.9599515176001212,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002428398058252427,
      "loss": 0.8169,
      "step": 2376
    },
    {
      "epoch": 0.9603555375991112,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002428155339805825,
      "loss": 0.885,
      "step": 2377
    },
    {
      "epoch": 0.960759557598101,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002427912621359223,
      "loss": 0.8568,
      "step": 2378
    },
    {
      "epoch": 0.961163577597091,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00024276699029126213,
      "loss": 0.9919,
      "step": 2379
    },
    {
      "epoch": 0.961567597596081,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002427427184466019,
      "loss": 0.7867,
      "step": 2380
    },
    {
      "epoch": 0.961971617595071,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00024271844660194174,
      "loss": 0.8673,
      "step": 2381
    },
    {
      "epoch": 0.9623756375940609,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00024269417475728154,
      "loss": 0.7885,
      "step": 2382
    },
    {
      "epoch": 0.9627796575930508,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00024266990291262131,
      "loss": 0.8356,
      "step": 2383
    },
    {
      "epoch": 0.9631836775920408,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00024264563106796114,
      "loss": 0.8174,
      "step": 2384
    },
    {
      "epoch": 0.9635876975910308,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00024262135922330095,
      "loss": 0.7922,
      "step": 2385
    },
    {
      "epoch": 0.9639917175900207,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00024259708737864075,
      "loss": 0.6862,
      "step": 2386
    },
    {
      "epoch": 0.9643957375890106,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00024257281553398058,
      "loss": 0.8054,
      "step": 2387
    },
    {
      "epoch": 0.9647997575880006,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00024254854368932035,
      "loss": 1.0563,
      "step": 2388
    },
    {
      "epoch": 0.9652037775869906,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00024252427184466018,
      "loss": 0.8065,
      "step": 2389
    },
    {
      "epoch": 0.9656077975859805,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00024249999999999999,
      "loss": 0.7753,
      "step": 2390
    },
    {
      "epoch": 0.9660118175849705,
      "grad_norm": 0.478515625,
      "learning_rate": 0.0002424757281553398,
      "loss": 0.8233,
      "step": 2391
    },
    {
      "epoch": 0.9664158375839604,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002424514563106796,
      "loss": 0.82,
      "step": 2392
    },
    {
      "epoch": 0.9668198575829503,
      "grad_norm": 0.625,
      "learning_rate": 0.0002424271844660194,
      "loss": 0.8846,
      "step": 2393
    },
    {
      "epoch": 0.9672238775819403,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002424029126213592,
      "loss": 0.7787,
      "step": 2394
    },
    {
      "epoch": 0.9676278975809303,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00024237864077669902,
      "loss": 0.8095,
      "step": 2395
    },
    {
      "epoch": 0.9680319175799202,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002423543689320388,
      "loss": 0.788,
      "step": 2396
    },
    {
      "epoch": 0.9684359375789101,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002423300970873786,
      "loss": 0.9105,
      "step": 2397
    },
    {
      "epoch": 0.9688399575779001,
      "grad_norm": 0.44921875,
      "learning_rate": 0.00024230582524271843,
      "loss": 0.7041,
      "step": 2398
    },
    {
      "epoch": 0.9692439775768901,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00024228155339805823,
      "loss": 0.76,
      "step": 2399
    },
    {
      "epoch": 0.96964799757588,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00024225728155339806,
      "loss": 0.7965,
      "step": 2400
    },
    {
      "epoch": 0.9700520175748699,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00024223300970873784,
      "loss": 0.8131,
      "step": 2401
    },
    {
      "epoch": 0.9704560375738599,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00024220873786407764,
      "loss": 0.7216,
      "step": 2402
    },
    {
      "epoch": 0.9708600575728499,
      "grad_norm": 0.53125,
      "learning_rate": 0.00024218446601941747,
      "loss": 0.8673,
      "step": 2403
    },
    {
      "epoch": 0.9712640775718399,
      "grad_norm": 0.5625,
      "learning_rate": 0.00024216019417475727,
      "loss": 0.8363,
      "step": 2404
    },
    {
      "epoch": 0.9716680975708297,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00024213592233009705,
      "loss": 0.8205,
      "step": 2405
    },
    {
      "epoch": 0.9720721175698197,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00024211165048543688,
      "loss": 0.8702,
      "step": 2406
    },
    {
      "epoch": 0.9724761375688097,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00024208737864077668,
      "loss": 0.7992,
      "step": 2407
    },
    {
      "epoch": 0.9728801575677996,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00024206310679611645,
      "loss": 0.8002,
      "step": 2408
    },
    {
      "epoch": 0.9732841775667895,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00024203883495145628,
      "loss": 0.7289,
      "step": 2409
    },
    {
      "epoch": 0.9736881975657795,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002420145631067961,
      "loss": 0.811,
      "step": 2410
    },
    {
      "epoch": 0.9740922175647695,
      "grad_norm": 0.53125,
      "learning_rate": 0.00024199029126213592,
      "loss": 0.7272,
      "step": 2411
    },
    {
      "epoch": 0.9744962375637594,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00024196601941747572,
      "loss": 0.8135,
      "step": 2412
    },
    {
      "epoch": 0.9749002575627493,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002419417475728155,
      "loss": 0.7297,
      "step": 2413
    },
    {
      "epoch": 0.9753042775617393,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00024191747572815532,
      "loss": 0.8167,
      "step": 2414
    },
    {
      "epoch": 0.9757082975607293,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00024189320388349513,
      "loss": 0.9756,
      "step": 2415
    },
    {
      "epoch": 0.9761123175597192,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00024186893203883493,
      "loss": 0.8452,
      "step": 2416
    },
    {
      "epoch": 0.9765163375587091,
      "grad_norm": 0.59375,
      "learning_rate": 0.00024184466019417476,
      "loss": 0.8521,
      "step": 2417
    },
    {
      "epoch": 0.9769203575576991,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00024182038834951453,
      "loss": 0.7248,
      "step": 2418
    },
    {
      "epoch": 0.9773243775566891,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00024179611650485433,
      "loss": 0.7601,
      "step": 2419
    },
    {
      "epoch": 0.977728397555679,
      "grad_norm": 0.546875,
      "learning_rate": 0.00024177184466019416,
      "loss": 0.8844,
      "step": 2420
    },
    {
      "epoch": 0.978132417554669,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00024174757281553394,
      "loss": 0.7862,
      "step": 2421
    },
    {
      "epoch": 0.9785364375536589,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00024172330097087377,
      "loss": 0.7872,
      "step": 2422
    },
    {
      "epoch": 0.9789404575526488,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00024169902912621357,
      "loss": 0.8859,
      "step": 2423
    },
    {
      "epoch": 0.9793444775516388,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00024167475728155337,
      "loss": 0.7984,
      "step": 2424
    },
    {
      "epoch": 0.9797484975506288,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0002416504854368932,
      "loss": 0.8413,
      "step": 2425
    },
    {
      "epoch": 0.9801525175496187,
      "grad_norm": 0.484375,
      "learning_rate": 0.00024162621359223298,
      "loss": 0.8328,
      "step": 2426
    },
    {
      "epoch": 0.9805565375486086,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00024160194174757278,
      "loss": 0.8012,
      "step": 2427
    },
    {
      "epoch": 0.9809605575475986,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002415776699029126,
      "loss": 0.7994,
      "step": 2428
    },
    {
      "epoch": 0.9813645775465886,
      "grad_norm": 0.44140625,
      "learning_rate": 0.0002415533980582524,
      "loss": 0.7204,
      "step": 2429
    },
    {
      "epoch": 0.9817685975455785,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00024152912621359221,
      "loss": 0.8335,
      "step": 2430
    },
    {
      "epoch": 0.9821726175445684,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00024150485436893202,
      "loss": 0.7426,
      "step": 2431
    },
    {
      "epoch": 0.9825766375435584,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00024148058252427182,
      "loss": 0.7487,
      "step": 2432
    },
    {
      "epoch": 0.9829806575425484,
      "grad_norm": 0.625,
      "learning_rate": 0.00024145631067961165,
      "loss": 0.8205,
      "step": 2433
    },
    {
      "epoch": 0.9833846775415384,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00024143203883495142,
      "loss": 0.9071,
      "step": 2434
    },
    {
      "epoch": 0.9837886975405282,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00024140776699029123,
      "loss": 0.9609,
      "step": 2435
    },
    {
      "epoch": 0.9841927175395182,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00024138349514563106,
      "loss": 0.8492,
      "step": 2436
    },
    {
      "epoch": 0.9845967375385082,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00024135922330097086,
      "loss": 0.8571,
      "step": 2437
    },
    {
      "epoch": 0.9850007575374982,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00024133495145631063,
      "loss": 0.8154,
      "step": 2438
    },
    {
      "epoch": 0.985404777536488,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00024131067961165046,
      "loss": 0.9073,
      "step": 2439
    },
    {
      "epoch": 0.985808797535478,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00024128640776699027,
      "loss": 0.823,
      "step": 2440
    },
    {
      "epoch": 0.986212817534468,
      "grad_norm": 0.44140625,
      "learning_rate": 0.0002412621359223301,
      "loss": 0.6553,
      "step": 2441
    },
    {
      "epoch": 0.9866168375334579,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002412378640776699,
      "loss": 0.8678,
      "step": 2442
    },
    {
      "epoch": 0.9870208575324478,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00024121359223300967,
      "loss": 0.8356,
      "step": 2443
    },
    {
      "epoch": 0.9874248775314378,
      "grad_norm": 0.462890625,
      "learning_rate": 0.0002411893203883495,
      "loss": 0.867,
      "step": 2444
    },
    {
      "epoch": 0.9878288975304278,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002411650485436893,
      "loss": 0.8268,
      "step": 2445
    },
    {
      "epoch": 0.9882329175294177,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00024114077669902908,
      "loss": 0.8013,
      "step": 2446
    },
    {
      "epoch": 0.9886369375284076,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002411165048543689,
      "loss": 0.9483,
      "step": 2447
    },
    {
      "epoch": 0.9890409575273976,
      "grad_norm": 0.5,
      "learning_rate": 0.0002410922330097087,
      "loss": 0.7918,
      "step": 2448
    },
    {
      "epoch": 0.9894449775263876,
      "grad_norm": 0.423828125,
      "learning_rate": 0.0002410679611650485,
      "loss": 0.7221,
      "step": 2449
    },
    {
      "epoch": 0.9898489975253775,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00024104368932038834,
      "loss": 0.7196,
      "step": 2450
    },
    {
      "epoch": 0.9902530175243675,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00024101941747572812,
      "loss": 0.8404,
      "step": 2451
    },
    {
      "epoch": 0.9906570375233574,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00024099514563106795,
      "loss": 0.8332,
      "step": 2452
    },
    {
      "epoch": 0.9910610575223474,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00024097087378640775,
      "loss": 0.8611,
      "step": 2453
    },
    {
      "epoch": 0.9914650775213373,
      "grad_norm": 0.515625,
      "learning_rate": 0.00024094660194174755,
      "loss": 0.7987,
      "step": 2454
    },
    {
      "epoch": 0.9918690975203273,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00024092233009708738,
      "loss": 0.8154,
      "step": 2455
    },
    {
      "epoch": 0.9922731175193172,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00024089805825242716,
      "loss": 0.8416,
      "step": 2456
    },
    {
      "epoch": 0.9926771375183071,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00024087378640776696,
      "loss": 0.9035,
      "step": 2457
    },
    {
      "epoch": 0.9930811575172971,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002408495145631068,
      "loss": 0.9502,
      "step": 2458
    },
    {
      "epoch": 0.9934851775162871,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00024082524271844656,
      "loss": 0.7709,
      "step": 2459
    },
    {
      "epoch": 0.993889197515277,
      "grad_norm": 0.5,
      "learning_rate": 0.0002408009708737864,
      "loss": 0.8336,
      "step": 2460
    },
    {
      "epoch": 0.9942932175142669,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002407766990291262,
      "loss": 0.8775,
      "step": 2461
    },
    {
      "epoch": 0.9946972375132569,
      "grad_norm": 0.5859375,
      "learning_rate": 0.000240752427184466,
      "loss": 0.8024,
      "step": 2462
    },
    {
      "epoch": 0.9951012575122469,
      "grad_norm": 0.5,
      "learning_rate": 0.00024072815533980583,
      "loss": 0.6933,
      "step": 2463
    },
    {
      "epoch": 0.9955052775112369,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002407038834951456,
      "loss": 0.802,
      "step": 2464
    },
    {
      "epoch": 0.9959092975102267,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002406796116504854,
      "loss": 0.9256,
      "step": 2465
    },
    {
      "epoch": 0.9963133175092167,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00024065533980582523,
      "loss": 0.9415,
      "step": 2466
    },
    {
      "epoch": 0.9967173375082067,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00024063106796116504,
      "loss": 0.8145,
      "step": 2467
    },
    {
      "epoch": 0.9971213575071967,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002406067961165048,
      "loss": 0.8682,
      "step": 2468
    },
    {
      "epoch": 0.9975253775061865,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00024058252427184464,
      "loss": 0.903,
      "step": 2469
    },
    {
      "epoch": 0.9979293975051765,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00024055825242718444,
      "loss": 0.9728,
      "step": 2470
    },
    {
      "epoch": 0.9983334175041665,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00024053398058252427,
      "loss": 0.7879,
      "step": 2471
    },
    {
      "epoch": 0.9987374375031564,
      "grad_norm": 0.484375,
      "learning_rate": 0.00024050970873786405,
      "loss": 0.7909,
      "step": 2472
    },
    {
      "epoch": 0.9991414575021463,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00024048543689320385,
      "loss": 0.7694,
      "step": 2473
    },
    {
      "epoch": 0.9995454775011363,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00024046116504854368,
      "loss": 0.9309,
      "step": 2474
    },
    {
      "epoch": 0.9999494975001263,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00024043689320388348,
      "loss": 0.8757,
      "step": 2475
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00024041262135922326,
      "loss": 0.6065,
      "step": 2476
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.582169771194458,
      "eval_runtime": 214.302,
      "eval_samples_per_second": 28.66,
      "eval_steps_per_second": 28.66,
      "step": 2476
    },
    {
      "epoch": 1.0,
      "eval_title2sid_loss": 1.17689049243927,
      "eval_title2sid_runtime": 136.7525,
      "eval_title2sid_samples_per_second": 27.831,
      "eval_title2sid_steps_per_second": 27.831,
      "step": 2476
    },
    {
      "epoch": 1.0,
      "eval_title2sid_loss": 1.17689049243927,
      "eval_title2sid_runtime": 136.7525,
      "eval_title2sid_samples_per_second": 27.831,
      "eval_title2sid_steps_per_second": 27.831,
      "step": 2476
    },
    {
      "epoch": 1.0,
      "eval_sid2title_loss": 0.3095063865184784,
      "eval_sid2title_runtime": 133.7967,
      "eval_sid2title_samples_per_second": 28.603,
      "eval_sid2title_steps_per_second": 28.603,
      "step": 2476
    },
    {
      "epoch": 1.0,
      "eval_sid2title_loss": 0.3095063865184784,
      "eval_sid2title_runtime": 133.7967,
      "eval_sid2title_samples_per_second": 28.603,
      "eval_sid2title_steps_per_second": 28.603,
      "step": 2476
    },
    {
      "epoch": 1.0004040199989899,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0002403883495145631,
      "loss": 0.7187,
      "step": 2477
    },
    {
      "epoch": 1.00080803999798,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002403640776699029,
      "loss": 0.642,
      "step": 2478
    },
    {
      "epoch": 1.0012120599969698,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002403398058252427,
      "loss": 0.677,
      "step": 2479
    },
    {
      "epoch": 1.0016160799959597,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00024031553398058252,
      "loss": 0.6586,
      "step": 2480
    },
    {
      "epoch": 1.0020200999949498,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0002402912621359223,
      "loss": 0.6067,
      "step": 2481
    },
    {
      "epoch": 1.0024241199939397,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00024026699029126213,
      "loss": 0.6873,
      "step": 2482
    },
    {
      "epoch": 1.0028281399929297,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00024024271844660193,
      "loss": 0.6084,
      "step": 2483
    },
    {
      "epoch": 1.0032321599919196,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0002402184466019417,
      "loss": 0.651,
      "step": 2484
    },
    {
      "epoch": 1.0036361799909095,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00024019417475728153,
      "loss": 0.715,
      "step": 2485
    },
    {
      "epoch": 1.0040401999898996,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00024016990291262134,
      "loss": 0.5067,
      "step": 2486
    },
    {
      "epoch": 1.0044442199888894,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00024014563106796114,
      "loss": 0.5855,
      "step": 2487
    },
    {
      "epoch": 1.0048482399878793,
      "grad_norm": 0.5,
      "learning_rate": 0.00024012135922330097,
      "loss": 0.6369,
      "step": 2488
    },
    {
      "epoch": 1.0052522599868694,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00024009708737864074,
      "loss": 0.6105,
      "step": 2489
    },
    {
      "epoch": 1.0056562799858593,
      "grad_norm": 0.65625,
      "learning_rate": 0.00024007281553398057,
      "loss": 0.6978,
      "step": 2490
    },
    {
      "epoch": 1.0060602999848494,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00024004854368932037,
      "loss": 0.6266,
      "step": 2491
    },
    {
      "epoch": 1.0064643199838392,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00024002427184466018,
      "loss": 0.6198,
      "step": 2492
    },
    {
      "epoch": 1.006868339982829,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00023999999999999998,
      "loss": 0.6427,
      "step": 2493
    },
    {
      "epoch": 1.0072723599818192,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00023997572815533978,
      "loss": 0.5797,
      "step": 2494
    },
    {
      "epoch": 1.007676379980809,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023995145631067958,
      "loss": 0.6279,
      "step": 2495
    },
    {
      "epoch": 1.008080399979799,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002399271844660194,
      "loss": 0.6687,
      "step": 2496
    },
    {
      "epoch": 1.008484419978789,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002399029126213592,
      "loss": 0.6112,
      "step": 2497
    },
    {
      "epoch": 1.0088884399777789,
      "grad_norm": 0.5625,
      "learning_rate": 0.000239878640776699,
      "loss": 0.6217,
      "step": 2498
    },
    {
      "epoch": 1.0092924599767688,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00023985436893203882,
      "loss": 0.6079,
      "step": 2499
    },
    {
      "epoch": 1.0096964799757588,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023983009708737862,
      "loss": 0.6188,
      "step": 2500
    },
    {
      "epoch": 1.0101004999747487,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00023980582524271845,
      "loss": 0.6205,
      "step": 2501
    },
    {
      "epoch": 1.0105045199737388,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00023978155339805823,
      "loss": 0.624,
      "step": 2502
    },
    {
      "epoch": 1.0109085399727287,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00023975728155339803,
      "loss": 0.6325,
      "step": 2503
    },
    {
      "epoch": 1.0113125599717185,
      "grad_norm": 0.53125,
      "learning_rate": 0.00023973300970873786,
      "loss": 0.6086,
      "step": 2504
    },
    {
      "epoch": 1.0117165799707086,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00023970873786407766,
      "loss": 0.6182,
      "step": 2505
    },
    {
      "epoch": 1.0121205999696985,
      "grad_norm": 0.53125,
      "learning_rate": 0.00023968446601941744,
      "loss": 0.6279,
      "step": 2506
    },
    {
      "epoch": 1.0125246199686884,
      "grad_norm": 0.578125,
      "learning_rate": 0.00023966019417475727,
      "loss": 0.6523,
      "step": 2507
    },
    {
      "epoch": 1.0129286399676785,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023963592233009707,
      "loss": 0.6298,
      "step": 2508
    },
    {
      "epoch": 1.0133326599666683,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023961165048543684,
      "loss": 0.6395,
      "step": 2509
    },
    {
      "epoch": 1.0137366799656582,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00023958737864077667,
      "loss": 0.6636,
      "step": 2510
    },
    {
      "epoch": 1.0141406999646483,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00023956310679611648,
      "loss": 0.6637,
      "step": 2511
    },
    {
      "epoch": 1.0145447199636382,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002395388349514563,
      "loss": 0.7063,
      "step": 2512
    },
    {
      "epoch": 1.0149487399626282,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002395145631067961,
      "loss": 0.6278,
      "step": 2513
    },
    {
      "epoch": 1.0153527599616181,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00023949029126213588,
      "loss": 0.6792,
      "step": 2514
    },
    {
      "epoch": 1.015756779960608,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002394660194174757,
      "loss": 0.5815,
      "step": 2515
    },
    {
      "epoch": 1.016160799959598,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00023944174757281551,
      "loss": 0.6637,
      "step": 2516
    },
    {
      "epoch": 1.016564819958588,
      "grad_norm": 0.431640625,
      "learning_rate": 0.00023941747572815532,
      "loss": 0.5477,
      "step": 2517
    },
    {
      "epoch": 1.0169688399575778,
      "grad_norm": 0.5,
      "learning_rate": 0.00023939320388349515,
      "loss": 0.5707,
      "step": 2518
    },
    {
      "epoch": 1.017372859956568,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00023936893203883492,
      "loss": 0.6742,
      "step": 2519
    },
    {
      "epoch": 1.0177768799555578,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00023934466019417472,
      "loss": 0.717,
      "step": 2520
    },
    {
      "epoch": 1.0181808999545479,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00023932038834951455,
      "loss": 0.6633,
      "step": 2521
    },
    {
      "epoch": 1.0185849199535377,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00023929611650485433,
      "loss": 0.6604,
      "step": 2522
    },
    {
      "epoch": 1.0189889399525276,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00023927184466019416,
      "loss": 0.6839,
      "step": 2523
    },
    {
      "epoch": 1.0193929599515177,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023924757281553396,
      "loss": 0.6571,
      "step": 2524
    },
    {
      "epoch": 1.0197969799505076,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00023922330097087376,
      "loss": 0.6464,
      "step": 2525
    },
    {
      "epoch": 1.0202009999494974,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0002391990291262136,
      "loss": 0.6756,
      "step": 2526
    },
    {
      "epoch": 1.0206050199484875,
      "grad_norm": 0.5625,
      "learning_rate": 0.00023917475728155337,
      "loss": 0.6615,
      "step": 2527
    },
    {
      "epoch": 1.0210090399474774,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00023915048543689317,
      "loss": 0.6513,
      "step": 2528
    },
    {
      "epoch": 1.0214130599464672,
      "grad_norm": 0.61328125,
      "learning_rate": 0.000239126213592233,
      "loss": 0.6333,
      "step": 2529
    },
    {
      "epoch": 1.0218170799454573,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002391019417475728,
      "loss": 0.596,
      "step": 2530
    },
    {
      "epoch": 1.0222210999444472,
      "grad_norm": 0.453125,
      "learning_rate": 0.0002390776699029126,
      "loss": 0.5896,
      "step": 2531
    },
    {
      "epoch": 1.0226251199434373,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002390533980582524,
      "loss": 0.7025,
      "step": 2532
    },
    {
      "epoch": 1.0230291399424272,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002390291262135922,
      "loss": 0.6333,
      "step": 2533
    },
    {
      "epoch": 1.023433159941417,
      "grad_norm": 0.609375,
      "learning_rate": 0.00023900485436893204,
      "loss": 0.6266,
      "step": 2534
    },
    {
      "epoch": 1.0238371799404071,
      "grad_norm": 0.458984375,
      "learning_rate": 0.0002389805825242718,
      "loss": 0.5724,
      "step": 2535
    },
    {
      "epoch": 1.024241199939397,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023895631067961162,
      "loss": 0.644,
      "step": 2536
    },
    {
      "epoch": 1.0246452199383869,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00023893203883495144,
      "loss": 0.6339,
      "step": 2537
    },
    {
      "epoch": 1.025049239937377,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00023890776699029125,
      "loss": 0.6635,
      "step": 2538
    },
    {
      "epoch": 1.0254532599363668,
      "grad_norm": 0.546875,
      "learning_rate": 0.00023888349514563102,
      "loss": 0.6459,
      "step": 2539
    },
    {
      "epoch": 1.0258572799353567,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00023885922330097085,
      "loss": 0.6353,
      "step": 2540
    },
    {
      "epoch": 1.0262612999343468,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023883495145631065,
      "loss": 0.7637,
      "step": 2541
    },
    {
      "epoch": 1.0266653199333367,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00023881067961165048,
      "loss": 0.6834,
      "step": 2542
    },
    {
      "epoch": 1.0270693399323267,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00023878640776699029,
      "loss": 0.6115,
      "step": 2543
    },
    {
      "epoch": 1.0274733599313166,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023876213592233006,
      "loss": 0.6358,
      "step": 2544
    },
    {
      "epoch": 1.0278773799303065,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002387378640776699,
      "loss": 0.7311,
      "step": 2545
    },
    {
      "epoch": 1.0282813999292966,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002387135922330097,
      "loss": 0.6402,
      "step": 2546
    },
    {
      "epoch": 1.0286854199282864,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023868932038834947,
      "loss": 0.6095,
      "step": 2547
    },
    {
      "epoch": 1.0290894399272763,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002386650485436893,
      "loss": 0.5385,
      "step": 2548
    },
    {
      "epoch": 1.0294934599262664,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002386407766990291,
      "loss": 0.636,
      "step": 2549
    },
    {
      "epoch": 1.0298974799252563,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002386165048543689,
      "loss": 0.6655,
      "step": 2550
    },
    {
      "epoch": 1.0303014999242464,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00023859223300970873,
      "loss": 0.7073,
      "step": 2551
    },
    {
      "epoch": 1.0307055199232362,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002385679611650485,
      "loss": 0.614,
      "step": 2552
    },
    {
      "epoch": 1.031109539922226,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00023854368932038834,
      "loss": 0.729,
      "step": 2553
    },
    {
      "epoch": 1.0315135599212162,
      "grad_norm": 0.59375,
      "learning_rate": 0.00023851941747572814,
      "loss": 0.6979,
      "step": 2554
    },
    {
      "epoch": 1.031917579920206,
      "grad_norm": 0.546875,
      "learning_rate": 0.00023849514563106794,
      "loss": 0.6647,
      "step": 2555
    },
    {
      "epoch": 1.032321599919196,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00023847087378640774,
      "loss": 0.8056,
      "step": 2556
    },
    {
      "epoch": 1.032725619918186,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00023844660194174755,
      "loss": 0.6144,
      "step": 2557
    },
    {
      "epoch": 1.0331296399171759,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00023842233009708735,
      "loss": 0.5566,
      "step": 2558
    },
    {
      "epoch": 1.0335336599161657,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00023839805825242718,
      "loss": 0.5884,
      "step": 2559
    },
    {
      "epoch": 1.0339376799151558,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00023837378640776695,
      "loss": 0.6065,
      "step": 2560
    },
    {
      "epoch": 1.0343416999141457,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00023834951456310678,
      "loss": 0.6614,
      "step": 2561
    },
    {
      "epoch": 1.0347457199131358,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00023832524271844658,
      "loss": 0.6355,
      "step": 2562
    },
    {
      "epoch": 1.0351497399121257,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002383009708737864,
      "loss": 0.6912,
      "step": 2563
    },
    {
      "epoch": 1.0355537599111155,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00023827669902912622,
      "loss": 0.5746,
      "step": 2564
    },
    {
      "epoch": 1.0359577799101056,
      "grad_norm": 0.51953125,
      "learning_rate": 0.000238252427184466,
      "loss": 0.6522,
      "step": 2565
    },
    {
      "epoch": 1.0363617999090955,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002382281553398058,
      "loss": 0.6347,
      "step": 2566
    },
    {
      "epoch": 1.0367658199080854,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00023820388349514562,
      "loss": 0.6004,
      "step": 2567
    },
    {
      "epoch": 1.0371698399070755,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00023817961165048543,
      "loss": 0.5656,
      "step": 2568
    },
    {
      "epoch": 1.0375738599060653,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002381553398058252,
      "loss": 0.6058,
      "step": 2569
    },
    {
      "epoch": 1.0379778799050552,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023813106796116503,
      "loss": 0.6526,
      "step": 2570
    },
    {
      "epoch": 1.0383818999040453,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00023810679611650483,
      "loss": 0.663,
      "step": 2571
    },
    {
      "epoch": 1.0387859199030351,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00023808252427184466,
      "loss": 0.5645,
      "step": 2572
    },
    {
      "epoch": 1.0391899399020252,
      "grad_norm": 0.515625,
      "learning_rate": 0.00023805825242718444,
      "loss": 0.6588,
      "step": 2573
    },
    {
      "epoch": 1.039593959901015,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00023803398058252424,
      "loss": 0.6525,
      "step": 2574
    },
    {
      "epoch": 1.039997979900005,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023800970873786407,
      "loss": 0.6371,
      "step": 2575
    },
    {
      "epoch": 1.040401999898995,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00023798543689320387,
      "loss": 0.6408,
      "step": 2576
    },
    {
      "epoch": 1.040806019897985,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00023796116504854365,
      "loss": 0.6807,
      "step": 2577
    },
    {
      "epoch": 1.0412100398969748,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00023793689320388348,
      "loss": 0.6741,
      "step": 2578
    },
    {
      "epoch": 1.041614059895965,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00023791262135922328,
      "loss": 0.6421,
      "step": 2579
    },
    {
      "epoch": 1.0420180798949548,
      "grad_norm": 0.5625,
      "learning_rate": 0.00023788834951456308,
      "loss": 0.6752,
      "step": 2580
    },
    {
      "epoch": 1.0424220998939449,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002378640776699029,
      "loss": 0.6381,
      "step": 2581
    },
    {
      "epoch": 1.0428261198929347,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023783980582524269,
      "loss": 0.5701,
      "step": 2582
    },
    {
      "epoch": 1.0432301398919246,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00023781553398058251,
      "loss": 0.6297,
      "step": 2583
    },
    {
      "epoch": 1.0436341598909147,
      "grad_norm": 0.703125,
      "learning_rate": 0.00023779126213592232,
      "loss": 0.7234,
      "step": 2584
    },
    {
      "epoch": 1.0440381798899045,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002377669902912621,
      "loss": 0.593,
      "step": 2585
    },
    {
      "epoch": 1.0444421998888944,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023774271844660192,
      "loss": 0.6959,
      "step": 2586
    },
    {
      "epoch": 1.0448462198878845,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023771844660194172,
      "loss": 0.6133,
      "step": 2587
    },
    {
      "epoch": 1.0452502398868744,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00023769417475728153,
      "loss": 0.6307,
      "step": 2588
    },
    {
      "epoch": 1.0456542598858642,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00023766990291262136,
      "loss": 0.5794,
      "step": 2589
    },
    {
      "epoch": 1.0460582798848543,
      "grad_norm": 0.75390625,
      "learning_rate": 0.00023764563106796113,
      "loss": 0.6444,
      "step": 2590
    },
    {
      "epoch": 1.0464622998838442,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00023762135922330096,
      "loss": 0.6009,
      "step": 2591
    },
    {
      "epoch": 1.0468663198828343,
      "grad_norm": 0.546875,
      "learning_rate": 0.00023759708737864076,
      "loss": 0.6135,
      "step": 2592
    },
    {
      "epoch": 1.0472703398818242,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00023757281553398057,
      "loss": 0.6476,
      "step": 2593
    },
    {
      "epoch": 1.047674359880814,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00023754854368932037,
      "loss": 0.6628,
      "step": 2594
    },
    {
      "epoch": 1.0480783798798041,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00023752427184466017,
      "loss": 0.6309,
      "step": 2595
    },
    {
      "epoch": 1.048482399878794,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023749999999999997,
      "loss": 0.6709,
      "step": 2596
    },
    {
      "epoch": 1.0488864198777839,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002374757281553398,
      "loss": 0.6564,
      "step": 2597
    },
    {
      "epoch": 1.049290439876774,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00023745145631067958,
      "loss": 0.5685,
      "step": 2598
    },
    {
      "epoch": 1.0496944598757638,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00023742718446601938,
      "loss": 0.6169,
      "step": 2599
    },
    {
      "epoch": 1.050098479874754,
      "grad_norm": 0.447265625,
      "learning_rate": 0.0002374029126213592,
      "loss": 0.6011,
      "step": 2600
    },
    {
      "epoch": 1.0505024998737438,
      "grad_norm": 0.6015625,
      "learning_rate": 0.000237378640776699,
      "loss": 0.6913,
      "step": 2601
    },
    {
      "epoch": 1.0509065198727336,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00023735436893203884,
      "loss": 0.5935,
      "step": 2602
    },
    {
      "epoch": 1.0513105398717237,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00023733009708737862,
      "loss": 0.5769,
      "step": 2603
    },
    {
      "epoch": 1.0517145598707136,
      "grad_norm": 0.75,
      "learning_rate": 0.00023730582524271842,
      "loss": 0.6944,
      "step": 2604
    },
    {
      "epoch": 1.0521185798697035,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00023728155339805825,
      "loss": 0.6961,
      "step": 2605
    },
    {
      "epoch": 1.0525225998686936,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00023725728155339805,
      "loss": 0.6464,
      "step": 2606
    },
    {
      "epoch": 1.0529266198676834,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00023723300970873783,
      "loss": 0.6172,
      "step": 2607
    },
    {
      "epoch": 1.0533306398666733,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023720873786407765,
      "loss": 0.6273,
      "step": 2608
    },
    {
      "epoch": 1.0537346598656634,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00023718446601941746,
      "loss": 0.6325,
      "step": 2609
    },
    {
      "epoch": 1.0541386798646533,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00023716019417475723,
      "loss": 0.5675,
      "step": 2610
    },
    {
      "epoch": 1.0545426998636434,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00023713592233009706,
      "loss": 0.6156,
      "step": 2611
    },
    {
      "epoch": 1.0549467198626332,
      "grad_norm": 0.515625,
      "learning_rate": 0.00023711165048543686,
      "loss": 0.6076,
      "step": 2612
    },
    {
      "epoch": 1.055350739861623,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002370873786407767,
      "loss": 0.6973,
      "step": 2613
    },
    {
      "epoch": 1.0557547598606132,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002370631067961165,
      "loss": 0.5916,
      "step": 2614
    },
    {
      "epoch": 1.056158779859603,
      "grad_norm": 0.53125,
      "learning_rate": 0.00023703883495145627,
      "loss": 0.601,
      "step": 2615
    },
    {
      "epoch": 1.056562799858593,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002370145631067961,
      "loss": 0.6077,
      "step": 2616
    },
    {
      "epoch": 1.056966819857583,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002369902912621359,
      "loss": 0.6046,
      "step": 2617
    },
    {
      "epoch": 1.0573708398565729,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002369660194174757,
      "loss": 0.6556,
      "step": 2618
    },
    {
      "epoch": 1.0577748598555627,
      "grad_norm": 0.609375,
      "learning_rate": 0.00023694174757281553,
      "loss": 0.6414,
      "step": 2619
    },
    {
      "epoch": 1.0581788798545528,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002369174757281553,
      "loss": 0.6256,
      "step": 2620
    },
    {
      "epoch": 1.0585828998535427,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002368932038834951,
      "loss": 0.6229,
      "step": 2621
    },
    {
      "epoch": 1.0589869198525328,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00023686893203883494,
      "loss": 0.629,
      "step": 2622
    },
    {
      "epoch": 1.0593909398515227,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00023684466019417472,
      "loss": 0.5684,
      "step": 2623
    },
    {
      "epoch": 1.0597949598505125,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00023682038834951455,
      "loss": 0.6184,
      "step": 2624
    },
    {
      "epoch": 1.0601989798495026,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00023679611650485435,
      "loss": 0.6569,
      "step": 2625
    },
    {
      "epoch": 1.0606029998484925,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023677184466019415,
      "loss": 0.6273,
      "step": 2626
    },
    {
      "epoch": 1.0610070198474824,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00023674757281553398,
      "loss": 0.6063,
      "step": 2627
    },
    {
      "epoch": 1.0614110398464724,
      "grad_norm": 0.53125,
      "learning_rate": 0.00023672330097087376,
      "loss": 0.591,
      "step": 2628
    },
    {
      "epoch": 1.0618150598454623,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00023669902912621356,
      "loss": 0.571,
      "step": 2629
    },
    {
      "epoch": 1.0622190798444524,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002366747572815534,
      "loss": 0.6736,
      "step": 2630
    },
    {
      "epoch": 1.0626230998434423,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002366504854368932,
      "loss": 0.6567,
      "step": 2631
    },
    {
      "epoch": 1.0630271198424321,
      "grad_norm": 0.625,
      "learning_rate": 0.000236626213592233,
      "loss": 0.6736,
      "step": 2632
    },
    {
      "epoch": 1.0634311398414222,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002366019417475728,
      "loss": 0.6682,
      "step": 2633
    },
    {
      "epoch": 1.063835159840412,
      "grad_norm": 0.4296875,
      "learning_rate": 0.0002365776699029126,
      "loss": 0.5849,
      "step": 2634
    },
    {
      "epoch": 1.064239179839402,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00023655339805825243,
      "loss": 0.598,
      "step": 2635
    },
    {
      "epoch": 1.064643199838392,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002365291262135922,
      "loss": 0.6548,
      "step": 2636
    },
    {
      "epoch": 1.065047219837382,
      "grad_norm": 0.671875,
      "learning_rate": 0.000236504854368932,
      "loss": 0.6975,
      "step": 2637
    },
    {
      "epoch": 1.0654512398363718,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00023648058252427183,
      "loss": 0.6769,
      "step": 2638
    },
    {
      "epoch": 1.065855259835362,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00023645631067961164,
      "loss": 0.6746,
      "step": 2639
    },
    {
      "epoch": 1.0662592798343518,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002364320388349514,
      "loss": 0.5758,
      "step": 2640
    },
    {
      "epoch": 1.0666632998333418,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00023640776699029124,
      "loss": 0.5712,
      "step": 2641
    },
    {
      "epoch": 1.0670673198323317,
      "grad_norm": 0.75,
      "learning_rate": 0.00023638349514563104,
      "loss": 0.7304,
      "step": 2642
    },
    {
      "epoch": 1.0674713398313216,
      "grad_norm": 0.65625,
      "learning_rate": 0.00023635922330097087,
      "loss": 0.6654,
      "step": 2643
    },
    {
      "epoch": 1.0678753598303117,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00023633495145631067,
      "loss": 0.7117,
      "step": 2644
    },
    {
      "epoch": 1.0682793798293015,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023631067961165045,
      "loss": 0.6652,
      "step": 2645
    },
    {
      "epoch": 1.0686833998282914,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023628640776699028,
      "loss": 0.8525,
      "step": 2646
    },
    {
      "epoch": 1.0690874198272815,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023626213592233008,
      "loss": 0.5976,
      "step": 2647
    },
    {
      "epoch": 1.0694914398262714,
      "grad_norm": 0.5,
      "learning_rate": 0.00023623786407766986,
      "loss": 0.567,
      "step": 2648
    },
    {
      "epoch": 1.0698954598252612,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023621359223300969,
      "loss": 0.6012,
      "step": 2649
    },
    {
      "epoch": 1.0702994798242513,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002361893203883495,
      "loss": 0.7015,
      "step": 2650
    },
    {
      "epoch": 1.0707034998232412,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002361650485436893,
      "loss": 0.6033,
      "step": 2651
    },
    {
      "epoch": 1.0711075198222313,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00023614077669902912,
      "loss": 0.6318,
      "step": 2652
    },
    {
      "epoch": 1.0715115398212212,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002361165048543689,
      "loss": 0.6655,
      "step": 2653
    },
    {
      "epoch": 1.071915559820211,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00023609223300970873,
      "loss": 0.5778,
      "step": 2654
    },
    {
      "epoch": 1.0723195798192011,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00023606796116504853,
      "loss": 0.5438,
      "step": 2655
    },
    {
      "epoch": 1.072723599818191,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00023604368932038833,
      "loss": 0.6356,
      "step": 2656
    },
    {
      "epoch": 1.0731276198171809,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023601941747572813,
      "loss": 0.6455,
      "step": 2657
    },
    {
      "epoch": 1.073531639816171,
      "grad_norm": 0.796875,
      "learning_rate": 0.00023599514563106793,
      "loss": 0.5962,
      "step": 2658
    },
    {
      "epoch": 1.0739356598151608,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00023597087378640774,
      "loss": 0.5791,
      "step": 2659
    },
    {
      "epoch": 1.074339679814151,
      "grad_norm": 0.546875,
      "learning_rate": 0.00023594660194174757,
      "loss": 0.5814,
      "step": 2660
    },
    {
      "epoch": 1.0747436998131408,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00023592233009708734,
      "loss": 0.6439,
      "step": 2661
    },
    {
      "epoch": 1.0751477198121306,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00023589805825242717,
      "loss": 0.645,
      "step": 2662
    },
    {
      "epoch": 1.0755517398111207,
      "grad_norm": 0.453125,
      "learning_rate": 0.00023587378640776697,
      "loss": 0.5496,
      "step": 2663
    },
    {
      "epoch": 1.0759557598101106,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00023584951456310678,
      "loss": 0.604,
      "step": 2664
    },
    {
      "epoch": 1.0763597798091005,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002358252427184466,
      "loss": 0.6551,
      "step": 2665
    },
    {
      "epoch": 1.0767637998080906,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023580097087378638,
      "loss": 0.7027,
      "step": 2666
    },
    {
      "epoch": 1.0771678198070804,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00023577669902912618,
      "loss": 0.6348,
      "step": 2667
    },
    {
      "epoch": 1.0775718398060703,
      "grad_norm": 0.6015625,
      "learning_rate": 0.000235752427184466,
      "loss": 0.6735,
      "step": 2668
    },
    {
      "epoch": 1.0779758598050604,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023572815533980581,
      "loss": 0.6474,
      "step": 2669
    },
    {
      "epoch": 1.0783798798040503,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002357038834951456,
      "loss": 0.6253,
      "step": 2670
    },
    {
      "epoch": 1.0787838998030403,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00023567961165048542,
      "loss": 0.6302,
      "step": 2671
    },
    {
      "epoch": 1.0791879198020302,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023565533980582522,
      "loss": 0.6045,
      "step": 2672
    },
    {
      "epoch": 1.07959193980102,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023563106796116505,
      "loss": 0.6319,
      "step": 2673
    },
    {
      "epoch": 1.0799959598000102,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023560679611650483,
      "loss": 0.677,
      "step": 2674
    },
    {
      "epoch": 1.080399979799,
      "grad_norm": 0.75390625,
      "learning_rate": 0.00023558252427184463,
      "loss": 0.7009,
      "step": 2675
    },
    {
      "epoch": 1.08080399979799,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00023555825242718446,
      "loss": 0.5265,
      "step": 2676
    },
    {
      "epoch": 1.08120801979698,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00023553398058252426,
      "loss": 0.5982,
      "step": 2677
    },
    {
      "epoch": 1.0816120397959699,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00023550970873786404,
      "loss": 0.6513,
      "step": 2678
    },
    {
      "epoch": 1.08201605979496,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00023548543689320386,
      "loss": 0.6917,
      "step": 2679
    },
    {
      "epoch": 1.0824200797939498,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00023546116504854367,
      "loss": 0.6639,
      "step": 2680
    },
    {
      "epoch": 1.0828240997929397,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00023543689320388347,
      "loss": 0.7033,
      "step": 2681
    },
    {
      "epoch": 1.0832281197919298,
      "grad_norm": 0.75,
      "learning_rate": 0.0002354126213592233,
      "loss": 0.7631,
      "step": 2682
    },
    {
      "epoch": 1.0836321397909197,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00023538834951456307,
      "loss": 0.6212,
      "step": 2683
    },
    {
      "epoch": 1.0840361597899095,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002353640776699029,
      "loss": 0.5849,
      "step": 2684
    },
    {
      "epoch": 1.0844401797888996,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002353398058252427,
      "loss": 0.6638,
      "step": 2685
    },
    {
      "epoch": 1.0848441997878895,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00023531553398058248,
      "loss": 0.7474,
      "step": 2686
    },
    {
      "epoch": 1.0852482197868794,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002352912621359223,
      "loss": 0.6747,
      "step": 2687
    },
    {
      "epoch": 1.0856522397858694,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002352669902912621,
      "loss": 0.6325,
      "step": 2688
    },
    {
      "epoch": 1.0860562597848593,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00023524271844660192,
      "loss": 0.5817,
      "step": 2689
    },
    {
      "epoch": 1.0864602797838494,
      "grad_norm": 0.515625,
      "learning_rate": 0.00023521844660194174,
      "loss": 0.663,
      "step": 2690
    },
    {
      "epoch": 1.0868642997828393,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00023519417475728152,
      "loss": 0.5802,
      "step": 2691
    },
    {
      "epoch": 1.0872683197818291,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00023516990291262135,
      "loss": 0.5979,
      "step": 2692
    },
    {
      "epoch": 1.0876723397808192,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00023514563106796115,
      "loss": 0.6811,
      "step": 2693
    },
    {
      "epoch": 1.088076359779809,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00023512135922330095,
      "loss": 0.6143,
      "step": 2694
    },
    {
      "epoch": 1.088480379778799,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00023509708737864076,
      "loss": 0.5903,
      "step": 2695
    },
    {
      "epoch": 1.088884399777789,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00023507281553398056,
      "loss": 0.6265,
      "step": 2696
    },
    {
      "epoch": 1.089288419776779,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00023504854368932036,
      "loss": 0.6929,
      "step": 2697
    },
    {
      "epoch": 1.089692439775769,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002350242718446602,
      "loss": 0.6229,
      "step": 2698
    },
    {
      "epoch": 1.0900964597747589,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023499999999999997,
      "loss": 0.6617,
      "step": 2699
    },
    {
      "epoch": 1.0905004797737488,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00023497572815533977,
      "loss": 0.6148,
      "step": 2700
    },
    {
      "epoch": 1.0909044997727388,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002349514563106796,
      "loss": 0.6203,
      "step": 2701
    },
    {
      "epoch": 1.0913085197717287,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002349271844660194,
      "loss": 0.6503,
      "step": 2702
    },
    {
      "epoch": 1.0917125397707186,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023490291262135923,
      "loss": 0.6631,
      "step": 2703
    },
    {
      "epoch": 1.0921165597697087,
      "grad_norm": 0.56640625,
      "learning_rate": 0.000234878640776699,
      "loss": 0.6746,
      "step": 2704
    },
    {
      "epoch": 1.0925205797686985,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002348543689320388,
      "loss": 0.5825,
      "step": 2705
    },
    {
      "epoch": 1.0929245997676884,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00023483009708737864,
      "loss": 0.6053,
      "step": 2706
    },
    {
      "epoch": 1.0933286197666785,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00023480582524271844,
      "loss": 0.6552,
      "step": 2707
    },
    {
      "epoch": 1.0937326397656684,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00023478155339805821,
      "loss": 0.6709,
      "step": 2708
    },
    {
      "epoch": 1.0941366597646582,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00023475728155339804,
      "loss": 0.6604,
      "step": 2709
    },
    {
      "epoch": 1.0945406797636483,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00023473300970873785,
      "loss": 0.6674,
      "step": 2710
    },
    {
      "epoch": 1.0949446997626382,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00023470873786407762,
      "loss": 0.7394,
      "step": 2711
    },
    {
      "epoch": 1.0953487197616283,
      "grad_norm": 0.451171875,
      "learning_rate": 0.00023468446601941745,
      "loss": 0.6232,
      "step": 2712
    },
    {
      "epoch": 1.0957527397606182,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00023466019417475725,
      "loss": 0.5744,
      "step": 2713
    },
    {
      "epoch": 1.096156759759608,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00023463592233009708,
      "loss": 0.5881,
      "step": 2714
    },
    {
      "epoch": 1.0965607797585981,
      "grad_norm": 0.671875,
      "learning_rate": 0.00023461165048543688,
      "loss": 0.5816,
      "step": 2715
    },
    {
      "epoch": 1.096964799757588,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00023458737864077666,
      "loss": 0.6,
      "step": 2716
    },
    {
      "epoch": 1.0973688197565779,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002345631067961165,
      "loss": 0.6562,
      "step": 2717
    },
    {
      "epoch": 1.097772839755568,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002345388349514563,
      "loss": 0.6439,
      "step": 2718
    },
    {
      "epoch": 1.0981768597545578,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002345145631067961,
      "loss": 0.6526,
      "step": 2719
    },
    {
      "epoch": 1.098580879753548,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002344902912621359,
      "loss": 0.6213,
      "step": 2720
    },
    {
      "epoch": 1.0989848997525378,
      "grad_norm": 0.5,
      "learning_rate": 0.0002344660194174757,
      "loss": 0.6009,
      "step": 2721
    },
    {
      "epoch": 1.0993889197515276,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002344417475728155,
      "loss": 0.6478,
      "step": 2722
    },
    {
      "epoch": 1.0997929397505177,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00023441747572815533,
      "loss": 0.5254,
      "step": 2723
    },
    {
      "epoch": 1.1001969597495076,
      "grad_norm": 0.625,
      "learning_rate": 0.0002343932038834951,
      "loss": 0.6446,
      "step": 2724
    },
    {
      "epoch": 1.1006009797484975,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00023436893203883494,
      "loss": 0.6376,
      "step": 2725
    },
    {
      "epoch": 1.1010049997474876,
      "grad_norm": 0.6875,
      "learning_rate": 0.00023434466019417474,
      "loss": 0.7427,
      "step": 2726
    },
    {
      "epoch": 1.1014090197464774,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00023432038834951454,
      "loss": 0.6072,
      "step": 2727
    },
    {
      "epoch": 1.1018130397454673,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00023429611650485437,
      "loss": 0.7211,
      "step": 2728
    },
    {
      "epoch": 1.1022170597444574,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00023427184466019414,
      "loss": 0.6538,
      "step": 2729
    },
    {
      "epoch": 1.1026210797434473,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00023424757281553395,
      "loss": 0.6543,
      "step": 2730
    },
    {
      "epoch": 1.1030250997424373,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00023422330097087378,
      "loss": 0.6772,
      "step": 2731
    },
    {
      "epoch": 1.1034291197414272,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00023419902912621358,
      "loss": 0.6434,
      "step": 2732
    },
    {
      "epoch": 1.103833139740417,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00023417475728155338,
      "loss": 0.5985,
      "step": 2733
    },
    {
      "epoch": 1.1042371597394072,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00023415048543689318,
      "loss": 0.5976,
      "step": 2734
    },
    {
      "epoch": 1.104641179738397,
      "grad_norm": 0.546875,
      "learning_rate": 0.00023412621359223299,
      "loss": 0.5814,
      "step": 2735
    },
    {
      "epoch": 1.105045199737387,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00023410194174757282,
      "loss": 0.6047,
      "step": 2736
    },
    {
      "epoch": 1.105449219736377,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002340776699029126,
      "loss": 0.5981,
      "step": 2737
    },
    {
      "epoch": 1.1058532397353669,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002340533980582524,
      "loss": 0.6162,
      "step": 2738
    },
    {
      "epoch": 1.106257259734357,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00023402912621359222,
      "loss": 0.6092,
      "step": 2739
    },
    {
      "epoch": 1.1066612797333468,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00023400485436893202,
      "loss": 0.5619,
      "step": 2740
    },
    {
      "epoch": 1.1070652997323367,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002339805825242718,
      "loss": 0.6803,
      "step": 2741
    },
    {
      "epoch": 1.1074693197313268,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00023395631067961163,
      "loss": 0.6158,
      "step": 2742
    },
    {
      "epoch": 1.1078733397303167,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00023393203883495143,
      "loss": 0.7064,
      "step": 2743
    },
    {
      "epoch": 1.1082773597293065,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00023390776699029126,
      "loss": 0.7374,
      "step": 2744
    },
    {
      "epoch": 1.1086813797282966,
      "grad_norm": 0.65625,
      "learning_rate": 0.00023388349514563106,
      "loss": 0.6024,
      "step": 2745
    },
    {
      "epoch": 1.1090853997272865,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00023385922330097084,
      "loss": 0.6616,
      "step": 2746
    },
    {
      "epoch": 1.1094894197262763,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00023383495145631067,
      "loss": 0.5919,
      "step": 2747
    },
    {
      "epoch": 1.1098934397252664,
      "grad_norm": 0.5,
      "learning_rate": 0.00023381067961165047,
      "loss": 0.631,
      "step": 2748
    },
    {
      "epoch": 1.1102974597242563,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00023378640776699025,
      "loss": 0.6313,
      "step": 2749
    },
    {
      "epoch": 1.1107014797232464,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00023376213592233007,
      "loss": 0.6669,
      "step": 2750
    },
    {
      "epoch": 1.1111054997222363,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00023373786407766988,
      "loss": 0.6212,
      "step": 2751
    },
    {
      "epoch": 1.1115095197212261,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00023371359223300968,
      "loss": 0.5539,
      "step": 2752
    },
    {
      "epoch": 1.1119135397202162,
      "grad_norm": 0.4453125,
      "learning_rate": 0.0002336893203883495,
      "loss": 0.5884,
      "step": 2753
    },
    {
      "epoch": 1.112317559719206,
      "grad_norm": 2.078125,
      "learning_rate": 0.00023366504854368928,
      "loss": 0.6446,
      "step": 2754
    },
    {
      "epoch": 1.112721579718196,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00023364077669902911,
      "loss": 0.6866,
      "step": 2755
    },
    {
      "epoch": 1.113125599717186,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00023361650485436892,
      "loss": 0.588,
      "step": 2756
    },
    {
      "epoch": 1.113529619716176,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00023359223300970872,
      "loss": 0.6638,
      "step": 2757
    },
    {
      "epoch": 1.113933639715166,
      "grad_norm": 0.73828125,
      "learning_rate": 0.00023356796116504852,
      "loss": 0.7065,
      "step": 2758
    },
    {
      "epoch": 1.1143376597141559,
      "grad_norm": 0.625,
      "learning_rate": 0.00023354368932038832,
      "loss": 0.7035,
      "step": 2759
    },
    {
      "epoch": 1.1147416797131457,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00023351941747572813,
      "loss": 0.6817,
      "step": 2760
    },
    {
      "epoch": 1.1151456997121358,
      "grad_norm": 0.578125,
      "learning_rate": 0.00023349514563106795,
      "loss": 0.6492,
      "step": 2761
    },
    {
      "epoch": 1.1155497197111257,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00023347087378640773,
      "loss": 0.6268,
      "step": 2762
    },
    {
      "epoch": 1.1159537397101156,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00023344660194174756,
      "loss": 0.6292,
      "step": 2763
    },
    {
      "epoch": 1.1163577597091057,
      "grad_norm": 0.75,
      "learning_rate": 0.00023342233009708736,
      "loss": 0.6126,
      "step": 2764
    },
    {
      "epoch": 1.1167617797080955,
      "grad_norm": 0.65625,
      "learning_rate": 0.00023339805825242716,
      "loss": 0.7697,
      "step": 2765
    },
    {
      "epoch": 1.1171657997070854,
      "grad_norm": 0.5859375,
      "learning_rate": 0.000233373786407767,
      "loss": 0.6449,
      "step": 2766
    },
    {
      "epoch": 1.1175698197060755,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00023334951456310677,
      "loss": 0.5851,
      "step": 2767
    },
    {
      "epoch": 1.1179738397050654,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00023332524271844657,
      "loss": 0.6431,
      "step": 2768
    },
    {
      "epoch": 1.1183778597040555,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002333009708737864,
      "loss": 0.6306,
      "step": 2769
    },
    {
      "epoch": 1.1187818797030453,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002332766990291262,
      "loss": 0.6649,
      "step": 2770
    },
    {
      "epoch": 1.1191858997020352,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00023325242718446598,
      "loss": 0.645,
      "step": 2771
    },
    {
      "epoch": 1.1195899197010253,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002332281553398058,
      "loss": 0.678,
      "step": 2772
    },
    {
      "epoch": 1.1199939397000152,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002332038834951456,
      "loss": 0.6105,
      "step": 2773
    },
    {
      "epoch": 1.120397959699005,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00023317961165048544,
      "loss": 0.6134,
      "step": 2774
    },
    {
      "epoch": 1.120801979697995,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00023315533980582521,
      "loss": 0.5235,
      "step": 2775
    },
    {
      "epoch": 1.121205999696985,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00023313106796116502,
      "loss": 0.6946,
      "step": 2776
    },
    {
      "epoch": 1.1216100196959748,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00023310679611650485,
      "loss": 0.5917,
      "step": 2777
    },
    {
      "epoch": 1.122014039694965,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00023308252427184465,
      "loss": 0.6687,
      "step": 2778
    },
    {
      "epoch": 1.1224180596939548,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00023305825242718442,
      "loss": 0.6973,
      "step": 2779
    },
    {
      "epoch": 1.122822079692945,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00023303398058252425,
      "loss": 0.6144,
      "step": 2780
    },
    {
      "epoch": 1.1232260996919348,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00023300970873786406,
      "loss": 0.649,
      "step": 2781
    },
    {
      "epoch": 1.1236301196909246,
      "grad_norm": 0.5,
      "learning_rate": 0.00023298543689320386,
      "loss": 0.6289,
      "step": 2782
    },
    {
      "epoch": 1.1240341396899147,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002329611650485437,
      "loss": 0.6036,
      "step": 2783
    },
    {
      "epoch": 1.1244381596889046,
      "grad_norm": 0.546875,
      "learning_rate": 0.00023293689320388346,
      "loss": 0.6513,
      "step": 2784
    },
    {
      "epoch": 1.1248421796878945,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002329126213592233,
      "loss": 0.7036,
      "step": 2785
    },
    {
      "epoch": 1.1252461996868846,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002328883495145631,
      "loss": 0.6231,
      "step": 2786
    },
    {
      "epoch": 1.1256502196858744,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00023286407766990287,
      "loss": 0.5893,
      "step": 2787
    },
    {
      "epoch": 1.1260542396848643,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002328398058252427,
      "loss": 0.6353,
      "step": 2788
    },
    {
      "epoch": 1.1264582596838544,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002328155339805825,
      "loss": 0.5902,
      "step": 2789
    },
    {
      "epoch": 1.1268622796828442,
      "grad_norm": 0.419921875,
      "learning_rate": 0.0002327912621359223,
      "loss": 0.5546,
      "step": 2790
    },
    {
      "epoch": 1.1272662996818343,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00023276699029126213,
      "loss": 0.622,
      "step": 2791
    },
    {
      "epoch": 1.1276703196808242,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002327427184466019,
      "loss": 0.6151,
      "step": 2792
    },
    {
      "epoch": 1.128074339679814,
      "grad_norm": 0.609375,
      "learning_rate": 0.00023271844660194174,
      "loss": 0.6075,
      "step": 2793
    },
    {
      "epoch": 1.1284783596788042,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00023269417475728154,
      "loss": 0.6576,
      "step": 2794
    },
    {
      "epoch": 1.128882379677794,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00023266990291262134,
      "loss": 0.6591,
      "step": 2795
    },
    {
      "epoch": 1.1292863996767841,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023264563106796115,
      "loss": 0.6269,
      "step": 2796
    },
    {
      "epoch": 1.129690419675774,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00023262135922330095,
      "loss": 0.6385,
      "step": 2797
    },
    {
      "epoch": 1.1300944396747639,
      "grad_norm": 0.515625,
      "learning_rate": 0.00023259708737864075,
      "loss": 0.6228,
      "step": 2798
    },
    {
      "epoch": 1.130498459673754,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00023257281553398058,
      "loss": 0.7343,
      "step": 2799
    },
    {
      "epoch": 1.1309024796727438,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00023254854368932035,
      "loss": 0.6461,
      "step": 2800
    },
    {
      "epoch": 1.1313064996717337,
      "grad_norm": 0.44140625,
      "learning_rate": 0.00023252427184466016,
      "loss": 0.5399,
      "step": 2801
    },
    {
      "epoch": 1.1317105196707238,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023249999999999999,
      "loss": 0.715,
      "step": 2802
    },
    {
      "epoch": 1.1321145396697136,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002324757281553398,
      "loss": 0.5644,
      "step": 2803
    },
    {
      "epoch": 1.1325185596687035,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00023245145631067962,
      "loss": 0.6418,
      "step": 2804
    },
    {
      "epoch": 1.1329225796676936,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002324271844660194,
      "loss": 0.6915,
      "step": 2805
    },
    {
      "epoch": 1.1333265996666835,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002324029126213592,
      "loss": 0.5982,
      "step": 2806
    },
    {
      "epoch": 1.1337306196656733,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023237864077669903,
      "loss": 0.654,
      "step": 2807
    },
    {
      "epoch": 1.1341346396646634,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00023235436893203883,
      "loss": 0.5894,
      "step": 2808
    },
    {
      "epoch": 1.1345386596636533,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002323300970873786,
      "loss": 0.6119,
      "step": 2809
    },
    {
      "epoch": 1.1349426796626434,
      "grad_norm": 0.765625,
      "learning_rate": 0.00023230582524271843,
      "loss": 0.566,
      "step": 2810
    },
    {
      "epoch": 1.1353466996616333,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00023228155339805823,
      "loss": 0.6108,
      "step": 2811
    },
    {
      "epoch": 1.1357507196606231,
      "grad_norm": 0.50390625,
      "learning_rate": 0.000232257281553398,
      "loss": 0.635,
      "step": 2812
    },
    {
      "epoch": 1.1361547396596132,
      "grad_norm": 0.5625,
      "learning_rate": 0.00023223300970873784,
      "loss": 0.6665,
      "step": 2813
    },
    {
      "epoch": 1.136558759658603,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00023220873786407764,
      "loss": 0.5492,
      "step": 2814
    },
    {
      "epoch": 1.136962779657593,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00023218446601941747,
      "loss": 0.645,
      "step": 2815
    },
    {
      "epoch": 1.137366799656583,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00023216019417475727,
      "loss": 0.5562,
      "step": 2816
    },
    {
      "epoch": 1.137770819655573,
      "grad_norm": 0.546875,
      "learning_rate": 0.00023213592233009705,
      "loss": 0.6224,
      "step": 2817
    },
    {
      "epoch": 1.138174839654563,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00023211165048543688,
      "loss": 0.6852,
      "step": 2818
    },
    {
      "epoch": 1.1385788596535529,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00023208737864077668,
      "loss": 0.6595,
      "step": 2819
    },
    {
      "epoch": 1.1389828796525427,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00023206310679611648,
      "loss": 0.6021,
      "step": 2820
    },
    {
      "epoch": 1.1393868996515328,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00023203883495145629,
      "loss": 0.5979,
      "step": 2821
    },
    {
      "epoch": 1.1397909196505227,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002320145631067961,
      "loss": 0.6003,
      "step": 2822
    },
    {
      "epoch": 1.1401949396495126,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002319902912621359,
      "loss": 0.7235,
      "step": 2823
    },
    {
      "epoch": 1.1405989596485027,
      "grad_norm": 0.53125,
      "learning_rate": 0.00023196601941747572,
      "loss": 0.6038,
      "step": 2824
    },
    {
      "epoch": 1.1410029796474925,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002319417475728155,
      "loss": 0.6534,
      "step": 2825
    },
    {
      "epoch": 1.1414069996464824,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00023191747572815532,
      "loss": 0.6298,
      "step": 2826
    },
    {
      "epoch": 1.1418110196454725,
      "grad_norm": 0.484375,
      "learning_rate": 0.00023189320388349513,
      "loss": 0.5727,
      "step": 2827
    },
    {
      "epoch": 1.1422150396444624,
      "grad_norm": 0.484375,
      "learning_rate": 0.00023186893203883493,
      "loss": 0.5984,
      "step": 2828
    },
    {
      "epoch": 1.1426190596434522,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00023184466019417476,
      "loss": 0.6497,
      "step": 2829
    },
    {
      "epoch": 1.1430230796424423,
      "grad_norm": 0.703125,
      "learning_rate": 0.00023182038834951453,
      "loss": 0.6806,
      "step": 2830
    },
    {
      "epoch": 1.1434270996414322,
      "grad_norm": 0.578125,
      "learning_rate": 0.00023179611650485434,
      "loss": 0.6242,
      "step": 2831
    },
    {
      "epoch": 1.1438311196404223,
      "grad_norm": 0.578125,
      "learning_rate": 0.00023177184466019417,
      "loss": 0.6558,
      "step": 2832
    },
    {
      "epoch": 1.1442351396394121,
      "grad_norm": 0.53125,
      "learning_rate": 0.00023174757281553397,
      "loss": 0.66,
      "step": 2833
    },
    {
      "epoch": 1.144639159638402,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00023172330097087377,
      "loss": 0.64,
      "step": 2834
    },
    {
      "epoch": 1.145043179637392,
      "grad_norm": 0.546875,
      "learning_rate": 0.00023169902912621357,
      "loss": 0.6199,
      "step": 2835
    },
    {
      "epoch": 1.145447199636382,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00023167475728155337,
      "loss": 0.6064,
      "step": 2836
    },
    {
      "epoch": 1.145851219635372,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0002316504854368932,
      "loss": 0.6279,
      "step": 2837
    },
    {
      "epoch": 1.146255239634362,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00023162621359223298,
      "loss": 0.5549,
      "step": 2838
    },
    {
      "epoch": 1.1466592596333518,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00023160194174757278,
      "loss": 0.6374,
      "step": 2839
    },
    {
      "epoch": 1.147063279632342,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002315776699029126,
      "loss": 0.6862,
      "step": 2840
    },
    {
      "epoch": 1.1474672996313318,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002315533980582524,
      "loss": 0.7036,
      "step": 2841
    },
    {
      "epoch": 1.1478713196303216,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002315291262135922,
      "loss": 0.6651,
      "step": 2842
    },
    {
      "epoch": 1.1482753396293117,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023150485436893202,
      "loss": 0.5972,
      "step": 2843
    },
    {
      "epoch": 1.1486793596283016,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00023148058252427182,
      "loss": 0.6264,
      "step": 2844
    },
    {
      "epoch": 1.1490833796272915,
      "grad_norm": 0.546875,
      "learning_rate": 0.00023145631067961165,
      "loss": 0.5288,
      "step": 2845
    },
    {
      "epoch": 1.1494873996262815,
      "grad_norm": 0.578125,
      "learning_rate": 0.00023143203883495145,
      "loss": 0.6815,
      "step": 2846
    },
    {
      "epoch": 1.1498914196252714,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00023140776699029123,
      "loss": 0.6442,
      "step": 2847
    },
    {
      "epoch": 1.1502954396242613,
      "grad_norm": 0.73828125,
      "learning_rate": 0.00023138349514563106,
      "loss": 0.7739,
      "step": 2848
    },
    {
      "epoch": 1.1506994596232514,
      "grad_norm": 0.625,
      "learning_rate": 0.00023135922330097086,
      "loss": 0.7192,
      "step": 2849
    },
    {
      "epoch": 1.1511034796222412,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00023133495145631063,
      "loss": 0.6611,
      "step": 2850
    },
    {
      "epoch": 1.1515074996212313,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00023131067961165046,
      "loss": 0.67,
      "step": 2851
    },
    {
      "epoch": 1.1519115196202212,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00023128640776699027,
      "loss": 0.6631,
      "step": 2852
    },
    {
      "epoch": 1.152315539619211,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00023126213592233007,
      "loss": 0.6729,
      "step": 2853
    },
    {
      "epoch": 1.1527195596182012,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002312378640776699,
      "loss": 0.5821,
      "step": 2854
    },
    {
      "epoch": 1.153123579617191,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00023121359223300967,
      "loss": 0.6267,
      "step": 2855
    },
    {
      "epoch": 1.1535275996161811,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002311893203883495,
      "loss": 0.5908,
      "step": 2856
    },
    {
      "epoch": 1.153931619615171,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002311650485436893,
      "loss": 0.7274,
      "step": 2857
    },
    {
      "epoch": 1.1543356396141609,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002311407766990291,
      "loss": 0.6444,
      "step": 2858
    },
    {
      "epoch": 1.154739659613151,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002311165048543689,
      "loss": 0.6522,
      "step": 2859
    },
    {
      "epoch": 1.1551436796121408,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002310922330097087,
      "loss": 0.6456,
      "step": 2860
    },
    {
      "epoch": 1.1555476996111307,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00023106796116504851,
      "loss": 0.7205,
      "step": 2861
    },
    {
      "epoch": 1.1559517196101208,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00023104368932038834,
      "loss": 0.639,
      "step": 2862
    },
    {
      "epoch": 1.1563557396091106,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00023101941747572812,
      "loss": 0.582,
      "step": 2863
    },
    {
      "epoch": 1.1567597596081005,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00023099514563106795,
      "loss": 0.5623,
      "step": 2864
    },
    {
      "epoch": 1.1571637796070906,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00023097087378640775,
      "loss": 0.6209,
      "step": 2865
    },
    {
      "epoch": 1.1575677996060805,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023094660194174755,
      "loss": 0.6129,
      "step": 2866
    },
    {
      "epoch": 1.1579718196050703,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023092233009708738,
      "loss": 0.6451,
      "step": 2867
    },
    {
      "epoch": 1.1583758396040604,
      "grad_norm": 0.5625,
      "learning_rate": 0.00023089805825242716,
      "loss": 0.6155,
      "step": 2868
    },
    {
      "epoch": 1.1587798596030503,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00023087378640776696,
      "loss": 0.6128,
      "step": 2869
    },
    {
      "epoch": 1.1591838796020404,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002308495145631068,
      "loss": 0.7287,
      "step": 2870
    },
    {
      "epoch": 1.1595878996010303,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002308252427184466,
      "loss": 0.6099,
      "step": 2871
    },
    {
      "epoch": 1.1599919196000201,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00023080097087378637,
      "loss": 0.5753,
      "step": 2872
    },
    {
      "epoch": 1.1603959395990102,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002307766990291262,
      "loss": 0.6352,
      "step": 2873
    },
    {
      "epoch": 1.160799959598,
      "grad_norm": 0.6015625,
      "learning_rate": 0.000230752427184466,
      "loss": 0.6263,
      "step": 2874
    },
    {
      "epoch": 1.16120397959699,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00023072815533980583,
      "loss": 0.6274,
      "step": 2875
    },
    {
      "epoch": 1.16160799959598,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002307038834951456,
      "loss": 0.6381,
      "step": 2876
    },
    {
      "epoch": 1.16201201959497,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002306796116504854,
      "loss": 0.621,
      "step": 2877
    },
    {
      "epoch": 1.16241603959396,
      "grad_norm": 0.609375,
      "learning_rate": 0.00023065533980582524,
      "loss": 0.7122,
      "step": 2878
    },
    {
      "epoch": 1.1628200595929499,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00023063106796116504,
      "loss": 0.6792,
      "step": 2879
    },
    {
      "epoch": 1.1632240795919397,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002306067961165048,
      "loss": 0.596,
      "step": 2880
    },
    {
      "epoch": 1.1636280995909298,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00023058252427184464,
      "loss": 0.6686,
      "step": 2881
    },
    {
      "epoch": 1.1640321195899197,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00023055825242718444,
      "loss": 0.6046,
      "step": 2882
    },
    {
      "epoch": 1.1644361395889096,
      "grad_norm": 0.453125,
      "learning_rate": 0.00023053398058252425,
      "loss": 0.5581,
      "step": 2883
    },
    {
      "epoch": 1.1648401595878997,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00023050970873786405,
      "loss": 0.6124,
      "step": 2884
    },
    {
      "epoch": 1.1652441795868895,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00023048543689320385,
      "loss": 0.6749,
      "step": 2885
    },
    {
      "epoch": 1.1656481995858794,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00023046116504854368,
      "loss": 0.6051,
      "step": 2886
    },
    {
      "epoch": 1.1660522195848695,
      "grad_norm": 0.6875,
      "learning_rate": 0.00023043689320388348,
      "loss": 0.6973,
      "step": 2887
    },
    {
      "epoch": 1.1664562395838594,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00023041262135922326,
      "loss": 0.6339,
      "step": 2888
    },
    {
      "epoch": 1.1668602595828494,
      "grad_norm": 0.46875,
      "learning_rate": 0.0002303883495145631,
      "loss": 0.5784,
      "step": 2889
    },
    {
      "epoch": 1.1672642795818393,
      "grad_norm": 0.4453125,
      "learning_rate": 0.0002303640776699029,
      "loss": 0.5779,
      "step": 2890
    },
    {
      "epoch": 1.1676682995808292,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002303398058252427,
      "loss": 0.6698,
      "step": 2891
    },
    {
      "epoch": 1.1680723195798193,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00023031553398058252,
      "loss": 0.6285,
      "step": 2892
    },
    {
      "epoch": 1.1684763395788091,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002302912621359223,
      "loss": 0.6145,
      "step": 2893
    },
    {
      "epoch": 1.168880359577799,
      "grad_norm": 0.439453125,
      "learning_rate": 0.00023026699029126213,
      "loss": 0.5725,
      "step": 2894
    },
    {
      "epoch": 1.169284379576789,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00023024271844660193,
      "loss": 0.6744,
      "step": 2895
    },
    {
      "epoch": 1.169688399575779,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00023021844660194173,
      "loss": 0.6562,
      "step": 2896
    },
    {
      "epoch": 1.170092419574769,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00023019417475728153,
      "loss": 0.7428,
      "step": 2897
    },
    {
      "epoch": 1.170496439573759,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00023016990291262134,
      "loss": 0.6207,
      "step": 2898
    },
    {
      "epoch": 1.1709004595727488,
      "grad_norm": 0.578125,
      "learning_rate": 0.00023014563106796114,
      "loss": 0.6556,
      "step": 2899
    },
    {
      "epoch": 1.1713044795717389,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00023012135922330097,
      "loss": 0.5926,
      "step": 2900
    },
    {
      "epoch": 1.1717084995707288,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00023009708737864074,
      "loss": 0.5746,
      "step": 2901
    },
    {
      "epoch": 1.1721125195697186,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00023007281553398055,
      "loss": 0.5622,
      "step": 2902
    },
    {
      "epoch": 1.1725165395687087,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00023004854368932038,
      "loss": 0.6172,
      "step": 2903
    },
    {
      "epoch": 1.1729205595676986,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00023002427184466018,
      "loss": 0.6778,
      "step": 2904
    },
    {
      "epoch": 1.1733245795666885,
      "grad_norm": 0.439453125,
      "learning_rate": 0.00023,
      "loss": 0.563,
      "step": 2905
    },
    {
      "epoch": 1.1737285995656785,
      "grad_norm": 0.484375,
      "learning_rate": 0.00022997572815533978,
      "loss": 0.6677,
      "step": 2906
    },
    {
      "epoch": 1.1741326195646684,
      "grad_norm": 0.515625,
      "learning_rate": 0.00022995145631067958,
      "loss": 0.6481,
      "step": 2907
    },
    {
      "epoch": 1.1745366395636583,
      "grad_norm": 0.53125,
      "learning_rate": 0.00022992718446601941,
      "loss": 0.6394,
      "step": 2908
    },
    {
      "epoch": 1.1749406595626484,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00022990291262135922,
      "loss": 0.659,
      "step": 2909
    },
    {
      "epoch": 1.1753446795616382,
      "grad_norm": 0.56640625,
      "learning_rate": 0.000229878640776699,
      "loss": 0.6071,
      "step": 2910
    },
    {
      "epoch": 1.1757486995606283,
      "grad_norm": 0.53125,
      "learning_rate": 0.00022985436893203882,
      "loss": 0.6982,
      "step": 2911
    },
    {
      "epoch": 1.1761527195596182,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00022983009708737862,
      "loss": 0.6374,
      "step": 2912
    },
    {
      "epoch": 1.176556739558608,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002298058252427184,
      "loss": 0.6705,
      "step": 2913
    },
    {
      "epoch": 1.1769607595575982,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022978155339805823,
      "loss": 0.6543,
      "step": 2914
    },
    {
      "epoch": 1.177364779556588,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00022975728155339803,
      "loss": 0.6798,
      "step": 2915
    },
    {
      "epoch": 1.1777687995555781,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022973300970873786,
      "loss": 0.6604,
      "step": 2916
    },
    {
      "epoch": 1.178172819554568,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00022970873786407766,
      "loss": 0.6012,
      "step": 2917
    },
    {
      "epoch": 1.1785768395535579,
      "grad_norm": 0.640625,
      "learning_rate": 0.00022968446601941744,
      "loss": 0.6577,
      "step": 2918
    },
    {
      "epoch": 1.178980859552548,
      "grad_norm": 0.6875,
      "learning_rate": 0.00022966019417475727,
      "loss": 0.7368,
      "step": 2919
    },
    {
      "epoch": 1.1793848795515378,
      "grad_norm": 0.515625,
      "learning_rate": 0.00022963592233009707,
      "loss": 0.6272,
      "step": 2920
    },
    {
      "epoch": 1.1797888995505277,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00022961165048543687,
      "loss": 0.5667,
      "step": 2921
    },
    {
      "epoch": 1.1801929195495178,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00022958737864077667,
      "loss": 0.6569,
      "step": 2922
    },
    {
      "epoch": 1.1805969395485076,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00022956310679611648,
      "loss": 0.6392,
      "step": 2923
    },
    {
      "epoch": 1.1810009595474975,
      "grad_norm": 0.46875,
      "learning_rate": 0.00022953883495145628,
      "loss": 0.6147,
      "step": 2924
    },
    {
      "epoch": 1.1814049795464876,
      "grad_norm": 0.435546875,
      "learning_rate": 0.0002295145631067961,
      "loss": 0.6043,
      "step": 2925
    },
    {
      "epoch": 1.1818089995454775,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00022949029126213588,
      "loss": 0.6775,
      "step": 2926
    },
    {
      "epoch": 1.1822130195444673,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002294660194174757,
      "loss": 0.6235,
      "step": 2927
    },
    {
      "epoch": 1.1826170395434574,
      "grad_norm": 0.59375,
      "learning_rate": 0.00022944174757281551,
      "loss": 0.5505,
      "step": 2928
    },
    {
      "epoch": 1.1830210595424473,
      "grad_norm": 0.546875,
      "learning_rate": 0.00022941747572815532,
      "loss": 0.6077,
      "step": 2929
    },
    {
      "epoch": 1.1834250795414374,
      "grad_norm": 0.53125,
      "learning_rate": 0.00022939320388349515,
      "loss": 0.6393,
      "step": 2930
    },
    {
      "epoch": 1.1838290995404273,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022936893203883492,
      "loss": 0.6851,
      "step": 2931
    },
    {
      "epoch": 1.1842331195394171,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00022934466019417472,
      "loss": 0.6048,
      "step": 2932
    },
    {
      "epoch": 1.1846371395384072,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00022932038834951455,
      "loss": 0.559,
      "step": 2933
    },
    {
      "epoch": 1.185041159537397,
      "grad_norm": 0.53125,
      "learning_rate": 0.00022929611650485436,
      "loss": 0.6645,
      "step": 2934
    },
    {
      "epoch": 1.1854451795363872,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00022927184466019416,
      "loss": 0.7475,
      "step": 2935
    },
    {
      "epoch": 1.185849199535377,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022924757281553396,
      "loss": 0.6278,
      "step": 2936
    },
    {
      "epoch": 1.186253219534367,
      "grad_norm": 0.546875,
      "learning_rate": 0.00022922330097087376,
      "loss": 0.6764,
      "step": 2937
    },
    {
      "epoch": 1.186657239533357,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002291990291262136,
      "loss": 0.6947,
      "step": 2938
    },
    {
      "epoch": 1.1870612595323469,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00022917475728155337,
      "loss": 0.7014,
      "step": 2939
    },
    {
      "epoch": 1.1874652795313367,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00022915048543689317,
      "loss": 0.6263,
      "step": 2940
    },
    {
      "epoch": 1.1878692995303268,
      "grad_norm": 0.5390625,
      "learning_rate": 0.000229126213592233,
      "loss": 0.6674,
      "step": 2941
    },
    {
      "epoch": 1.1882733195293167,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002291019417475728,
      "loss": 0.7065,
      "step": 2942
    },
    {
      "epoch": 1.1886773395283066,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00022907766990291258,
      "loss": 0.6183,
      "step": 2943
    },
    {
      "epoch": 1.1890813595272967,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002290533980582524,
      "loss": 0.5799,
      "step": 2944
    },
    {
      "epoch": 1.1894853795262865,
      "grad_norm": 0.48046875,
      "learning_rate": 0.0002290291262135922,
      "loss": 0.6661,
      "step": 2945
    },
    {
      "epoch": 1.1898893995252764,
      "grad_norm": 0.5,
      "learning_rate": 0.00022900485436893204,
      "loss": 0.5795,
      "step": 2946
    },
    {
      "epoch": 1.1902934195242665,
      "grad_norm": 0.5625,
      "learning_rate": 0.00022898058252427184,
      "loss": 0.6358,
      "step": 2947
    },
    {
      "epoch": 1.1906974395232564,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00022895631067961162,
      "loss": 0.6328,
      "step": 2948
    },
    {
      "epoch": 1.1911014595222464,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00022893203883495145,
      "loss": 0.6236,
      "step": 2949
    },
    {
      "epoch": 1.1915054795212363,
      "grad_norm": 0.546875,
      "learning_rate": 0.00022890776699029125,
      "loss": 0.668,
      "step": 2950
    },
    {
      "epoch": 1.1919094995202262,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00022888349514563102,
      "loss": 0.619,
      "step": 2951
    },
    {
      "epoch": 1.1923135195192163,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00022885922330097085,
      "loss": 0.6701,
      "step": 2952
    },
    {
      "epoch": 1.1927175395182061,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00022883495145631065,
      "loss": 0.6196,
      "step": 2953
    },
    {
      "epoch": 1.193121559517196,
      "grad_norm": 0.625,
      "learning_rate": 0.00022881067961165046,
      "loss": 0.6951,
      "step": 2954
    },
    {
      "epoch": 1.193525579516186,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002287864077669903,
      "loss": 0.6031,
      "step": 2955
    },
    {
      "epoch": 1.193929599515176,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022876213592233006,
      "loss": 0.6037,
      "step": 2956
    },
    {
      "epoch": 1.194333619514166,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002287378640776699,
      "loss": 0.6188,
      "step": 2957
    },
    {
      "epoch": 1.194737639513156,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002287135922330097,
      "loss": 0.6582,
      "step": 2958
    },
    {
      "epoch": 1.1951416595121458,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002286893203883495,
      "loss": 0.6337,
      "step": 2959
    },
    {
      "epoch": 1.1955456795111359,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002286650485436893,
      "loss": 0.625,
      "step": 2960
    },
    {
      "epoch": 1.1959496995101258,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002286407766990291,
      "loss": 0.6396,
      "step": 2961
    },
    {
      "epoch": 1.1963537195091156,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0002286165048543689,
      "loss": 0.7611,
      "step": 2962
    },
    {
      "epoch": 1.1967577395081057,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00022859223300970873,
      "loss": 0.5095,
      "step": 2963
    },
    {
      "epoch": 1.1971617595070956,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002285679611650485,
      "loss": 0.6981,
      "step": 2964
    },
    {
      "epoch": 1.1975657795060854,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00022854368932038834,
      "loss": 0.6595,
      "step": 2965
    },
    {
      "epoch": 1.1979697995050755,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022851941747572814,
      "loss": 0.623,
      "step": 2966
    },
    {
      "epoch": 1.1983738195040654,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00022849514563106794,
      "loss": 0.6796,
      "step": 2967
    },
    {
      "epoch": 1.1987778395030555,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00022847087378640777,
      "loss": 0.6776,
      "step": 2968
    },
    {
      "epoch": 1.1991818595020454,
      "grad_norm": 0.546875,
      "learning_rate": 0.00022844660194174755,
      "loss": 0.5996,
      "step": 2969
    },
    {
      "epoch": 1.1995858795010352,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00022842233009708735,
      "loss": 0.5918,
      "step": 2970
    },
    {
      "epoch": 1.1999898995000253,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00022839805825242718,
      "loss": 0.6616,
      "step": 2971
    },
    {
      "epoch": 1.2003939194990152,
      "grad_norm": 0.53125,
      "learning_rate": 0.00022837378640776698,
      "loss": 0.5747,
      "step": 2972
    },
    {
      "epoch": 1.200797939498005,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00022834951456310676,
      "loss": 0.697,
      "step": 2973
    },
    {
      "epoch": 1.2012019594969952,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00022832524271844659,
      "loss": 0.6607,
      "step": 2974
    },
    {
      "epoch": 1.201605979495985,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002283009708737864,
      "loss": 0.6211,
      "step": 2975
    },
    {
      "epoch": 1.2020099994949751,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00022827669902912622,
      "loss": 0.5532,
      "step": 2976
    },
    {
      "epoch": 1.202414019493965,
      "grad_norm": 0.53515625,
      "learning_rate": 0.000228252427184466,
      "loss": 0.6515,
      "step": 2977
    },
    {
      "epoch": 1.2028180394929548,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002282281553398058,
      "loss": 0.6655,
      "step": 2978
    },
    {
      "epoch": 1.203222059491945,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00022820388349514562,
      "loss": 0.7328,
      "step": 2979
    },
    {
      "epoch": 1.2036260794909348,
      "grad_norm": 0.44140625,
      "learning_rate": 0.00022817961165048543,
      "loss": 0.5316,
      "step": 2980
    },
    {
      "epoch": 1.2040300994899247,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002281553398058252,
      "loss": 0.5576,
      "step": 2981
    },
    {
      "epoch": 1.2044341194889148,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00022813106796116503,
      "loss": 0.6336,
      "step": 2982
    },
    {
      "epoch": 1.2048381394879046,
      "grad_norm": 0.640625,
      "learning_rate": 0.00022810679611650483,
      "loss": 0.6814,
      "step": 2983
    },
    {
      "epoch": 1.2052421594868945,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00022808252427184464,
      "loss": 0.5864,
      "step": 2984
    },
    {
      "epoch": 1.2056461794858846,
      "grad_norm": 0.5625,
      "learning_rate": 0.00022805825242718444,
      "loss": 0.5819,
      "step": 2985
    },
    {
      "epoch": 1.2060501994848745,
      "grad_norm": 0.546875,
      "learning_rate": 0.00022803398058252424,
      "loss": 0.5951,
      "step": 2986
    },
    {
      "epoch": 1.2064542194838643,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022800970873786407,
      "loss": 0.6458,
      "step": 2987
    },
    {
      "epoch": 1.2068582394828544,
      "grad_norm": 0.515625,
      "learning_rate": 0.00022798543689320387,
      "loss": 0.611,
      "step": 2988
    },
    {
      "epoch": 1.2072622594818443,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022796116504854365,
      "loss": 0.6777,
      "step": 2989
    },
    {
      "epoch": 1.2076662794808344,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00022793689320388348,
      "loss": 0.5651,
      "step": 2990
    },
    {
      "epoch": 1.2080702994798242,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00022791262135922328,
      "loss": 0.711,
      "step": 2991
    },
    {
      "epoch": 1.2084743194788141,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00022788834951456308,
      "loss": 0.5651,
      "step": 2992
    },
    {
      "epoch": 1.2088783394778042,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002278640776699029,
      "loss": 0.6674,
      "step": 2993
    },
    {
      "epoch": 1.209282359476794,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00022783980582524269,
      "loss": 0.6407,
      "step": 2994
    },
    {
      "epoch": 1.2096863794757842,
      "grad_norm": 0.453125,
      "learning_rate": 0.00022781553398058252,
      "loss": 0.6424,
      "step": 2995
    },
    {
      "epoch": 1.210090399474774,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00022779126213592232,
      "loss": 0.5,
      "step": 2996
    },
    {
      "epoch": 1.210494419473764,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00022776699029126212,
      "loss": 0.5823,
      "step": 2997
    },
    {
      "epoch": 1.210898439472754,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00022774271844660192,
      "loss": 0.579,
      "step": 2998
    },
    {
      "epoch": 1.2113024594717439,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00022771844660194173,
      "loss": 0.6524,
      "step": 2999
    },
    {
      "epoch": 1.2117064794707337,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00022769417475728153,
      "loss": 0.6072,
      "step": 3000
    },
    {
      "epoch": 1.2121104994697238,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00022766990291262136,
      "loss": 0.6015,
      "step": 3001
    },
    {
      "epoch": 1.2125145194687137,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00022764563106796113,
      "loss": 0.5741,
      "step": 3002
    },
    {
      "epoch": 1.2129185394677036,
      "grad_norm": 0.5625,
      "learning_rate": 0.00022762135922330093,
      "loss": 0.6582,
      "step": 3003
    },
    {
      "epoch": 1.2133225594666937,
      "grad_norm": 0.4375,
      "learning_rate": 0.00022759708737864076,
      "loss": 0.6581,
      "step": 3004
    },
    {
      "epoch": 1.2137265794656835,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00022757281553398057,
      "loss": 0.6706,
      "step": 3005
    },
    {
      "epoch": 1.2141305994646734,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002275485436893204,
      "loss": 0.621,
      "step": 3006
    },
    {
      "epoch": 1.2145346194636635,
      "grad_norm": 0.515625,
      "learning_rate": 0.00022752427184466017,
      "loss": 0.5977,
      "step": 3007
    },
    {
      "epoch": 1.2149386394626533,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00022749999999999997,
      "loss": 0.5926,
      "step": 3008
    },
    {
      "epoch": 1.2153426594616434,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002274757281553398,
      "loss": 0.5955,
      "step": 3009
    },
    {
      "epoch": 1.2157466794606333,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002274514563106796,
      "loss": 0.6293,
      "step": 3010
    },
    {
      "epoch": 1.2161506994596232,
      "grad_norm": 0.578125,
      "learning_rate": 0.00022742718446601938,
      "loss": 0.6916,
      "step": 3011
    },
    {
      "epoch": 1.2165547194586133,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002274029126213592,
      "loss": 0.6747,
      "step": 3012
    },
    {
      "epoch": 1.2169587394576031,
      "grad_norm": 0.4921875,
      "learning_rate": 0.000227378640776699,
      "loss": 0.6004,
      "step": 3013
    },
    {
      "epoch": 1.2173627594565932,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002273543689320388,
      "loss": 0.5784,
      "step": 3014
    },
    {
      "epoch": 1.217766779455583,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00022733009708737862,
      "loss": 0.6585,
      "step": 3015
    },
    {
      "epoch": 1.218170799454573,
      "grad_norm": 0.5,
      "learning_rate": 0.00022730582524271842,
      "loss": 0.6723,
      "step": 3016
    },
    {
      "epoch": 1.218574819453563,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00022728155339805825,
      "loss": 0.6445,
      "step": 3017
    },
    {
      "epoch": 1.218978839452553,
      "grad_norm": 0.451171875,
      "learning_rate": 0.00022725728155339805,
      "loss": 0.5757,
      "step": 3018
    },
    {
      "epoch": 1.2193828594515428,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00022723300970873783,
      "loss": 0.6881,
      "step": 3019
    },
    {
      "epoch": 1.2197868794505329,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00022720873786407766,
      "loss": 0.6793,
      "step": 3020
    },
    {
      "epoch": 1.2201908994495227,
      "grad_norm": 0.546875,
      "learning_rate": 0.00022718446601941746,
      "loss": 0.6039,
      "step": 3021
    },
    {
      "epoch": 1.2205949194485126,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022716019417475726,
      "loss": 0.6879,
      "step": 3022
    },
    {
      "epoch": 1.2209989394475027,
      "grad_norm": 0.453125,
      "learning_rate": 0.00022713592233009706,
      "loss": 0.5525,
      "step": 3023
    },
    {
      "epoch": 1.2214029594464926,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00022711165048543686,
      "loss": 0.6323,
      "step": 3024
    },
    {
      "epoch": 1.2218069794454824,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00022708737864077667,
      "loss": 0.6766,
      "step": 3025
    },
    {
      "epoch": 1.2222109994444725,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0002270631067961165,
      "loss": 0.6601,
      "step": 3026
    },
    {
      "epoch": 1.2226150194434624,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00022703883495145627,
      "loss": 0.6738,
      "step": 3027
    },
    {
      "epoch": 1.2230190394424525,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002270145631067961,
      "loss": 0.6353,
      "step": 3028
    },
    {
      "epoch": 1.2234230594414424,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002269902912621359,
      "loss": 0.5983,
      "step": 3029
    },
    {
      "epoch": 1.2238270794404322,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0002269660194174757,
      "loss": 0.635,
      "step": 3030
    },
    {
      "epoch": 1.2242310994394223,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00022694174757281554,
      "loss": 0.6961,
      "step": 3031
    },
    {
      "epoch": 1.2246351194384122,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002269174757281553,
      "loss": 0.6178,
      "step": 3032
    },
    {
      "epoch": 1.225039139437402,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002268932038834951,
      "loss": 0.6007,
      "step": 3033
    },
    {
      "epoch": 1.2254431594363921,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022686893203883494,
      "loss": 0.5725,
      "step": 3034
    },
    {
      "epoch": 1.225847179435382,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00022684466019417474,
      "loss": 0.63,
      "step": 3035
    },
    {
      "epoch": 1.226251199434372,
      "grad_norm": 0.46875,
      "learning_rate": 0.00022682038834951455,
      "loss": 0.5602,
      "step": 3036
    },
    {
      "epoch": 1.226655219433362,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00022679611650485435,
      "loss": 0.6953,
      "step": 3037
    },
    {
      "epoch": 1.2270592394323518,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00022677184466019415,
      "loss": 0.6006,
      "step": 3038
    },
    {
      "epoch": 1.227463259431342,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00022674757281553398,
      "loss": 0.6344,
      "step": 3039
    },
    {
      "epoch": 1.2278672794303318,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00022672330097087376,
      "loss": 0.6611,
      "step": 3040
    },
    {
      "epoch": 1.2282712994293217,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00022669902912621356,
      "loss": 0.5959,
      "step": 3041
    },
    {
      "epoch": 1.2286753194283118,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0002266747572815534,
      "loss": 0.5891,
      "step": 3042
    },
    {
      "epoch": 1.2290793394273016,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002266504854368932,
      "loss": 0.6764,
      "step": 3043
    },
    {
      "epoch": 1.2294833594262915,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00022662621359223297,
      "loss": 0.6652,
      "step": 3044
    },
    {
      "epoch": 1.2298873794252816,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002266019417475728,
      "loss": 0.5992,
      "step": 3045
    },
    {
      "epoch": 1.2302913994242715,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002265776699029126,
      "loss": 0.634,
      "step": 3046
    },
    {
      "epoch": 1.2306954194232613,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022655339805825243,
      "loss": 0.6697,
      "step": 3047
    },
    {
      "epoch": 1.2310994394222514,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002265291262135922,
      "loss": 0.6413,
      "step": 3048
    },
    {
      "epoch": 1.2315034594212413,
      "grad_norm": 0.5078125,
      "learning_rate": 0.000226504854368932,
      "loss": 0.5343,
      "step": 3049
    },
    {
      "epoch": 1.2319074794202314,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00022648058252427183,
      "loss": 0.6335,
      "step": 3050
    },
    {
      "epoch": 1.2323114994192212,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00022645631067961164,
      "loss": 0.638,
      "step": 3051
    },
    {
      "epoch": 1.2327155194182111,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002264320388349514,
      "loss": 0.6709,
      "step": 3052
    },
    {
      "epoch": 1.2331195394172012,
      "grad_norm": 0.53125,
      "learning_rate": 0.00022640776699029124,
      "loss": 0.6678,
      "step": 3053
    },
    {
      "epoch": 1.233523559416191,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00022638349514563104,
      "loss": 0.5775,
      "step": 3054
    },
    {
      "epoch": 1.2339275794151812,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00022635922330097085,
      "loss": 0.6152,
      "step": 3055
    },
    {
      "epoch": 1.234331599414171,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022633495145631068,
      "loss": 0.6511,
      "step": 3056
    },
    {
      "epoch": 1.234735619413161,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00022631067961165045,
      "loss": 0.6967,
      "step": 3057
    },
    {
      "epoch": 1.235139639412151,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00022628640776699028,
      "loss": 0.6,
      "step": 3058
    },
    {
      "epoch": 1.2355436594111409,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00022626213592233008,
      "loss": 0.6334,
      "step": 3059
    },
    {
      "epoch": 1.2359476794101307,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00022623786407766988,
      "loss": 0.5561,
      "step": 3060
    },
    {
      "epoch": 1.2363516994091208,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002262135922330097,
      "loss": 0.6726,
      "step": 3061
    },
    {
      "epoch": 1.2367557194081107,
      "grad_norm": 0.5,
      "learning_rate": 0.0002261893203883495,
      "loss": 0.5927,
      "step": 3062
    },
    {
      "epoch": 1.2371597394071006,
      "grad_norm": 0.427734375,
      "learning_rate": 0.0002261650485436893,
      "loss": 0.5507,
      "step": 3063
    },
    {
      "epoch": 1.2375637594060906,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00022614077669902912,
      "loss": 0.5499,
      "step": 3064
    },
    {
      "epoch": 1.2379677794050805,
      "grad_norm": 0.443359375,
      "learning_rate": 0.0002261165048543689,
      "loss": 0.6028,
      "step": 3065
    },
    {
      "epoch": 1.2383717994040704,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00022609223300970873,
      "loss": 0.6535,
      "step": 3066
    },
    {
      "epoch": 1.2387758194030605,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022606796116504853,
      "loss": 0.6215,
      "step": 3067
    },
    {
      "epoch": 1.2391798394020503,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00022604368932038833,
      "loss": 0.6988,
      "step": 3068
    },
    {
      "epoch": 1.2395838594010404,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00022601941747572816,
      "loss": 0.8085,
      "step": 3069
    },
    {
      "epoch": 1.2399878794000303,
      "grad_norm": 0.515625,
      "learning_rate": 0.00022599514563106794,
      "loss": 0.6163,
      "step": 3070
    },
    {
      "epoch": 1.2403918993990202,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00022597087378640774,
      "loss": 0.5696,
      "step": 3071
    },
    {
      "epoch": 1.2407959193980103,
      "grad_norm": 0.65625,
      "learning_rate": 0.00022594660194174757,
      "loss": 0.6489,
      "step": 3072
    },
    {
      "epoch": 1.2411999393970001,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00022592233009708737,
      "loss": 0.663,
      "step": 3073
    },
    {
      "epoch": 1.2416039593959902,
      "grad_norm": 0.578125,
      "learning_rate": 0.00022589805825242714,
      "loss": 0.6576,
      "step": 3074
    },
    {
      "epoch": 1.24200797939498,
      "grad_norm": 0.578125,
      "learning_rate": 0.00022587378640776697,
      "loss": 0.6759,
      "step": 3075
    },
    {
      "epoch": 1.24241199939397,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00022584951456310678,
      "loss": 0.7291,
      "step": 3076
    },
    {
      "epoch": 1.24281601939296,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002258252427184466,
      "loss": 0.5859,
      "step": 3077
    },
    {
      "epoch": 1.24322003939195,
      "grad_norm": 0.703125,
      "learning_rate": 0.00022580097087378638,
      "loss": 0.7104,
      "step": 3078
    },
    {
      "epoch": 1.2436240593909398,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00022577669902912618,
      "loss": 0.6344,
      "step": 3079
    },
    {
      "epoch": 1.2440280793899299,
      "grad_norm": 0.4765625,
      "learning_rate": 0.000225752427184466,
      "loss": 0.5929,
      "step": 3080
    },
    {
      "epoch": 1.2444320993889197,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00022572815533980582,
      "loss": 0.6556,
      "step": 3081
    },
    {
      "epoch": 1.2448361193879096,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002257038834951456,
      "loss": 0.6159,
      "step": 3082
    },
    {
      "epoch": 1.2452401393868997,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00022567961165048542,
      "loss": 0.6185,
      "step": 3083
    },
    {
      "epoch": 1.2456441593858896,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00022565533980582522,
      "loss": 0.5967,
      "step": 3084
    },
    {
      "epoch": 1.2460481793848794,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00022563106796116502,
      "loss": 0.6162,
      "step": 3085
    },
    {
      "epoch": 1.2464521993838695,
      "grad_norm": 0.546875,
      "learning_rate": 0.00022560679611650483,
      "loss": 0.6039,
      "step": 3086
    },
    {
      "epoch": 1.2468562193828594,
      "grad_norm": 0.625,
      "learning_rate": 0.00022558252427184463,
      "loss": 0.7004,
      "step": 3087
    },
    {
      "epoch": 1.2472602393818495,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00022555825242718446,
      "loss": 0.6924,
      "step": 3088
    },
    {
      "epoch": 1.2476642593808394,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00022553398058252426,
      "loss": 0.6197,
      "step": 3089
    },
    {
      "epoch": 1.2480682793798292,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00022550970873786404,
      "loss": 0.6292,
      "step": 3090
    },
    {
      "epoch": 1.2484722993788193,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00022548543689320387,
      "loss": 0.5759,
      "step": 3091
    },
    {
      "epoch": 1.2488763193778092,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00022546116504854367,
      "loss": 0.681,
      "step": 3092
    },
    {
      "epoch": 1.2492803393767993,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022543689320388347,
      "loss": 0.6294,
      "step": 3093
    },
    {
      "epoch": 1.2496843593757891,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002254126213592233,
      "loss": 0.6519,
      "step": 3094
    },
    {
      "epoch": 1.250088379374779,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00022538834951456308,
      "loss": 0.658,
      "step": 3095
    },
    {
      "epoch": 1.250492399373769,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002253640776699029,
      "loss": 0.6487,
      "step": 3096
    },
    {
      "epoch": 1.250896419372759,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002253398058252427,
      "loss": 0.5974,
      "step": 3097
    },
    {
      "epoch": 1.2513004393717488,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002253155339805825,
      "loss": 0.6589,
      "step": 3098
    },
    {
      "epoch": 1.251704459370739,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002252912621359223,
      "loss": 0.695,
      "step": 3099
    },
    {
      "epoch": 1.2521084793697288,
      "grad_norm": 0.5625,
      "learning_rate": 0.00022526699029126211,
      "loss": 0.6701,
      "step": 3100
    },
    {
      "epoch": 1.2525124993687187,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022524271844660192,
      "loss": 0.6291,
      "step": 3101
    },
    {
      "epoch": 1.2529165193677088,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022521844660194175,
      "loss": 0.5957,
      "step": 3102
    },
    {
      "epoch": 1.2533205393666986,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022519417475728152,
      "loss": 0.5809,
      "step": 3103
    },
    {
      "epoch": 1.2537245593656885,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00022516990291262132,
      "loss": 0.6652,
      "step": 3104
    },
    {
      "epoch": 1.2541285793646786,
      "grad_norm": 0.5625,
      "learning_rate": 0.00022514563106796115,
      "loss": 0.6433,
      "step": 3105
    },
    {
      "epoch": 1.2545325993636685,
      "grad_norm": 0.65625,
      "learning_rate": 0.00022512135922330096,
      "loss": 0.6678,
      "step": 3106
    },
    {
      "epoch": 1.2549366193626583,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00022509708737864078,
      "loss": 0.5966,
      "step": 3107
    },
    {
      "epoch": 1.2553406393616484,
      "grad_norm": 0.65625,
      "learning_rate": 0.00022507281553398056,
      "loss": 0.7103,
      "step": 3108
    },
    {
      "epoch": 1.2557446593606383,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00022504854368932036,
      "loss": 0.7071,
      "step": 3109
    },
    {
      "epoch": 1.2561486793596284,
      "grad_norm": 0.42578125,
      "learning_rate": 0.0002250242718446602,
      "loss": 0.5828,
      "step": 3110
    },
    {
      "epoch": 1.2565526993586182,
      "grad_norm": 0.5234375,
      "learning_rate": 0.000225,
      "loss": 0.5748,
      "step": 3111
    },
    {
      "epoch": 1.2569567193576083,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00022497572815533977,
      "loss": 0.5992,
      "step": 3112
    },
    {
      "epoch": 1.2573607393565982,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002249514563106796,
      "loss": 0.6387,
      "step": 3113
    },
    {
      "epoch": 1.257764759355588,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002249271844660194,
      "loss": 0.6773,
      "step": 3114
    },
    {
      "epoch": 1.2581687793545782,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00022490291262135918,
      "loss": 0.6217,
      "step": 3115
    },
    {
      "epoch": 1.258572799353568,
      "grad_norm": 0.5078125,
      "learning_rate": 0.000224878640776699,
      "loss": 0.6751,
      "step": 3116
    },
    {
      "epoch": 1.258976819352558,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002248543689320388,
      "loss": 0.6621,
      "step": 3117
    },
    {
      "epoch": 1.259380839351548,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00022483009708737864,
      "loss": 0.6292,
      "step": 3118
    },
    {
      "epoch": 1.2597848593505379,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022480582524271844,
      "loss": 0.618,
      "step": 3119
    },
    {
      "epoch": 1.2601888793495277,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00022478155339805821,
      "loss": 0.6144,
      "step": 3120
    },
    {
      "epoch": 1.2605928993485178,
      "grad_norm": 0.53125,
      "learning_rate": 0.00022475728155339804,
      "loss": 0.6348,
      "step": 3121
    },
    {
      "epoch": 1.2609969193475077,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00022473300970873785,
      "loss": 0.6406,
      "step": 3122
    },
    {
      "epoch": 1.2614009393464976,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00022470873786407765,
      "loss": 0.5564,
      "step": 3123
    },
    {
      "epoch": 1.2618049593454876,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00022468446601941745,
      "loss": 0.5946,
      "step": 3124
    },
    {
      "epoch": 1.2622089793444775,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00022466019417475725,
      "loss": 0.6334,
      "step": 3125
    },
    {
      "epoch": 1.2626129993434674,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00022463592233009706,
      "loss": 0.6369,
      "step": 3126
    },
    {
      "epoch": 1.2630170193424575,
      "grad_norm": 0.578125,
      "learning_rate": 0.00022461165048543689,
      "loss": 0.6136,
      "step": 3127
    },
    {
      "epoch": 1.2634210393414473,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00022458737864077666,
      "loss": 0.5752,
      "step": 3128
    },
    {
      "epoch": 1.2638250593404374,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002245631067961165,
      "loss": 0.6628,
      "step": 3129
    },
    {
      "epoch": 1.2642290793394273,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002245388349514563,
      "loss": 0.7376,
      "step": 3130
    },
    {
      "epoch": 1.2646330993384174,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002245145631067961,
      "loss": 0.7014,
      "step": 3131
    },
    {
      "epoch": 1.2650371193374073,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00022449029126213592,
      "loss": 0.6512,
      "step": 3132
    },
    {
      "epoch": 1.2654411393363971,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002244660194174757,
      "loss": 0.6708,
      "step": 3133
    },
    {
      "epoch": 1.2658451593353872,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002244417475728155,
      "loss": 0.6739,
      "step": 3134
    },
    {
      "epoch": 1.266249179334377,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00022441747572815533,
      "loss": 0.7016,
      "step": 3135
    },
    {
      "epoch": 1.266653199333367,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00022439320388349513,
      "loss": 0.6151,
      "step": 3136
    },
    {
      "epoch": 1.267057219332357,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00022436893203883494,
      "loss": 0.6715,
      "step": 3137
    },
    {
      "epoch": 1.267461239331347,
      "grad_norm": 0.515625,
      "learning_rate": 0.00022434466019417474,
      "loss": 0.6529,
      "step": 3138
    },
    {
      "epoch": 1.2678652593303368,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00022432038834951454,
      "loss": 0.6186,
      "step": 3139
    },
    {
      "epoch": 1.2682692793293269,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00022429611650485437,
      "loss": 0.69,
      "step": 3140
    },
    {
      "epoch": 1.2686732993283167,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00022427184466019415,
      "loss": 0.5891,
      "step": 3141
    },
    {
      "epoch": 1.2690773193273066,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00022424757281553395,
      "loss": 0.6042,
      "step": 3142
    },
    {
      "epoch": 1.2694813393262967,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00022422330097087378,
      "loss": 0.534,
      "step": 3143
    },
    {
      "epoch": 1.2698853593252866,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00022419902912621358,
      "loss": 0.7062,
      "step": 3144
    },
    {
      "epoch": 1.2702893793242764,
      "grad_norm": 0.46875,
      "learning_rate": 0.00022417475728155335,
      "loss": 0.5998,
      "step": 3145
    },
    {
      "epoch": 1.2706933993232665,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00022415048543689318,
      "loss": 0.7113,
      "step": 3146
    },
    {
      "epoch": 1.2710974193222564,
      "grad_norm": 0.474609375,
      "learning_rate": 0.000224126213592233,
      "loss": 0.6454,
      "step": 3147
    },
    {
      "epoch": 1.2715014393212465,
      "grad_norm": 0.6875,
      "learning_rate": 0.00022410194174757282,
      "loss": 0.6604,
      "step": 3148
    },
    {
      "epoch": 1.2719054593202364,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002240776699029126,
      "loss": 0.7094,
      "step": 3149
    },
    {
      "epoch": 1.2723094793192262,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002240533980582524,
      "loss": 0.5945,
      "step": 3150
    },
    {
      "epoch": 1.2727134993182163,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00022402912621359222,
      "loss": 0.6893,
      "step": 3151
    },
    {
      "epoch": 1.2731175193172062,
      "grad_norm": 0.59375,
      "learning_rate": 0.00022400485436893203,
      "loss": 0.6204,
      "step": 3152
    },
    {
      "epoch": 1.2735215393161963,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002239805825242718,
      "loss": 0.6447,
      "step": 3153
    },
    {
      "epoch": 1.2739255593151861,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00022395631067961163,
      "loss": 0.6172,
      "step": 3154
    },
    {
      "epoch": 1.274329579314176,
      "grad_norm": 0.546875,
      "learning_rate": 0.00022393203883495143,
      "loss": 0.6606,
      "step": 3155
    },
    {
      "epoch": 1.274733599313166,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022390776699029123,
      "loss": 0.5303,
      "step": 3156
    },
    {
      "epoch": 1.275137619312156,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022388349514563106,
      "loss": 0.6354,
      "step": 3157
    },
    {
      "epoch": 1.2755416393111458,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00022385922330097084,
      "loss": 0.6351,
      "step": 3158
    },
    {
      "epoch": 1.275945659310136,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00022383495145631067,
      "loss": 0.6659,
      "step": 3159
    },
    {
      "epoch": 1.2763496793091258,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00022381067961165047,
      "loss": 0.6879,
      "step": 3160
    },
    {
      "epoch": 1.2767536993081157,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00022378640776699027,
      "loss": 0.6865,
      "step": 3161
    },
    {
      "epoch": 1.2771577193071058,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00022376213592233008,
      "loss": 0.5823,
      "step": 3162
    },
    {
      "epoch": 1.2775617393060956,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00022373786407766988,
      "loss": 0.6159,
      "step": 3163
    },
    {
      "epoch": 1.2779657593050855,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00022371359223300968,
      "loss": 0.6944,
      "step": 3164
    },
    {
      "epoch": 1.2783697793040756,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002236893203883495,
      "loss": 0.6033,
      "step": 3165
    },
    {
      "epoch": 1.2787737993030655,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00022366504854368929,
      "loss": 0.651,
      "step": 3166
    },
    {
      "epoch": 1.2791778193020553,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022364077669902911,
      "loss": 0.6198,
      "step": 3167
    },
    {
      "epoch": 1.2795818393010454,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00022361650485436892,
      "loss": 0.6544,
      "step": 3168
    },
    {
      "epoch": 1.2799858593000353,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00022359223300970872,
      "loss": 0.5368,
      "step": 3169
    },
    {
      "epoch": 1.2803898792990254,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00022356796116504855,
      "loss": 0.6065,
      "step": 3170
    },
    {
      "epoch": 1.2807938992980152,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00022354368932038832,
      "loss": 0.6685,
      "step": 3171
    },
    {
      "epoch": 1.2811979192970053,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00022351941747572813,
      "loss": 0.6893,
      "step": 3172
    },
    {
      "epoch": 1.2816019392959952,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00022349514563106796,
      "loss": 0.6508,
      "step": 3173
    },
    {
      "epoch": 1.282005959294985,
      "grad_norm": 0.625,
      "learning_rate": 0.00022347087378640776,
      "loss": 0.6408,
      "step": 3174
    },
    {
      "epoch": 1.2824099792939752,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00022344660194174753,
      "loss": 0.6613,
      "step": 3175
    },
    {
      "epoch": 1.282813999292965,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022342233009708736,
      "loss": 0.6147,
      "step": 3176
    },
    {
      "epoch": 1.283218019291955,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022339805825242717,
      "loss": 0.6672,
      "step": 3177
    },
    {
      "epoch": 1.283622039290945,
      "grad_norm": 0.71875,
      "learning_rate": 0.000223373786407767,
      "loss": 0.8321,
      "step": 3178
    },
    {
      "epoch": 1.2840260592899349,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00022334951456310677,
      "loss": 0.5539,
      "step": 3179
    },
    {
      "epoch": 1.2844300792889247,
      "grad_norm": 0.5625,
      "learning_rate": 0.00022332524271844657,
      "loss": 0.6721,
      "step": 3180
    },
    {
      "epoch": 1.2848340992879148,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002233009708737864,
      "loss": 0.6923,
      "step": 3181
    },
    {
      "epoch": 1.2852381192869047,
      "grad_norm": 0.6875,
      "learning_rate": 0.0002232766990291262,
      "loss": 0.6967,
      "step": 3182
    },
    {
      "epoch": 1.2856421392858945,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00022325242718446598,
      "loss": 0.6017,
      "step": 3183
    },
    {
      "epoch": 1.2860461592848846,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002232281553398058,
      "loss": 0.6246,
      "step": 3184
    },
    {
      "epoch": 1.2864501792838745,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002232038834951456,
      "loss": 0.6025,
      "step": 3185
    },
    {
      "epoch": 1.2868541992828644,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002231796116504854,
      "loss": 0.6774,
      "step": 3186
    },
    {
      "epoch": 1.2872582192818545,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00022315533980582522,
      "loss": 0.6025,
      "step": 3187
    },
    {
      "epoch": 1.2876622392808443,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00022313106796116502,
      "loss": 0.5127,
      "step": 3188
    },
    {
      "epoch": 1.2880662592798344,
      "grad_norm": 0.484375,
      "learning_rate": 0.00022310679611650485,
      "loss": 0.6149,
      "step": 3189
    },
    {
      "epoch": 1.2884702792788243,
      "grad_norm": 0.453125,
      "learning_rate": 0.00022308252427184465,
      "loss": 0.592,
      "step": 3190
    },
    {
      "epoch": 1.2888742992778144,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00022305825242718443,
      "loss": 0.6005,
      "step": 3191
    },
    {
      "epoch": 1.2892783192768043,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00022303398058252425,
      "loss": 0.5909,
      "step": 3192
    },
    {
      "epoch": 1.2896823392757941,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00022300970873786406,
      "loss": 0.7299,
      "step": 3193
    },
    {
      "epoch": 1.2900863592747842,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00022298543689320386,
      "loss": 0.5746,
      "step": 3194
    },
    {
      "epoch": 1.290490379273774,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002229611650485437,
      "loss": 0.6323,
      "step": 3195
    },
    {
      "epoch": 1.290894399272764,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00022293689320388346,
      "loss": 0.6349,
      "step": 3196
    },
    {
      "epoch": 1.291298419271754,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002229126213592233,
      "loss": 0.5936,
      "step": 3197
    },
    {
      "epoch": 1.291702439270744,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002228883495145631,
      "loss": 0.569,
      "step": 3198
    },
    {
      "epoch": 1.2921064592697338,
      "grad_norm": 0.625,
      "learning_rate": 0.0002228640776699029,
      "loss": 0.6698,
      "step": 3199
    },
    {
      "epoch": 1.2925104792687239,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0002228398058252427,
      "loss": 0.7301,
      "step": 3200
    },
    {
      "epoch": 1.2929144992677137,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002228155339805825,
      "loss": 0.6626,
      "step": 3201
    },
    {
      "epoch": 1.2933185192667036,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002227912621359223,
      "loss": 0.6999,
      "step": 3202
    },
    {
      "epoch": 1.2937225392656937,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00022276699029126213,
      "loss": 0.6062,
      "step": 3203
    },
    {
      "epoch": 1.2941265592646836,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0002227427184466019,
      "loss": 0.6995,
      "step": 3204
    },
    {
      "epoch": 1.2945305792636734,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002227184466019417,
      "loss": 0.6171,
      "step": 3205
    },
    {
      "epoch": 1.2949345992626635,
      "grad_norm": 0.515625,
      "learning_rate": 0.00022269417475728154,
      "loss": 0.6382,
      "step": 3206
    },
    {
      "epoch": 1.2953386192616534,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00022266990291262134,
      "loss": 0.6154,
      "step": 3207
    },
    {
      "epoch": 1.2957426392606435,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00022264563106796117,
      "loss": 0.6279,
      "step": 3208
    },
    {
      "epoch": 1.2961466592596333,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00022262135922330095,
      "loss": 0.6338,
      "step": 3209
    },
    {
      "epoch": 1.2965506792586234,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022259708737864075,
      "loss": 0.6783,
      "step": 3210
    },
    {
      "epoch": 1.2969546992576133,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00022257281553398058,
      "loss": 0.6556,
      "step": 3211
    },
    {
      "epoch": 1.2973587192566032,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00022254854368932036,
      "loss": 0.7159,
      "step": 3212
    },
    {
      "epoch": 1.2977627392555933,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00022252427184466016,
      "loss": 0.6505,
      "step": 3213
    },
    {
      "epoch": 1.2981667592545831,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002225,
      "loss": 0.6939,
      "step": 3214
    },
    {
      "epoch": 1.298570779253573,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002224757281553398,
      "loss": 0.6395,
      "step": 3215
    },
    {
      "epoch": 1.298974799252563,
      "grad_norm": 0.484375,
      "learning_rate": 0.00022245145631067956,
      "loss": 0.6565,
      "step": 3216
    },
    {
      "epoch": 1.299378819251553,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002224271844660194,
      "loss": 0.6171,
      "step": 3217
    },
    {
      "epoch": 1.2997828392505428,
      "grad_norm": 0.42578125,
      "learning_rate": 0.0002224029126213592,
      "loss": 0.5633,
      "step": 3218
    },
    {
      "epoch": 1.300186859249533,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00022237864077669903,
      "loss": 0.7106,
      "step": 3219
    },
    {
      "epoch": 1.3005908792485228,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00022235436893203883,
      "loss": 0.6676,
      "step": 3220
    },
    {
      "epoch": 1.3009948992475127,
      "grad_norm": 0.48046875,
      "learning_rate": 0.0002223300970873786,
      "loss": 0.5903,
      "step": 3221
    },
    {
      "epoch": 1.3013989192465028,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00022230582524271843,
      "loss": 0.6586,
      "step": 3222
    },
    {
      "epoch": 1.3018029392454926,
      "grad_norm": 0.5625,
      "learning_rate": 0.00022228155339805824,
      "loss": 0.6452,
      "step": 3223
    },
    {
      "epoch": 1.3022069592444825,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00022225728155339804,
      "loss": 0.6641,
      "step": 3224
    },
    {
      "epoch": 1.3026109792434726,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00022223300970873784,
      "loss": 0.6946,
      "step": 3225
    },
    {
      "epoch": 1.3030149992424624,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00022220873786407764,
      "loss": 0.5461,
      "step": 3226
    },
    {
      "epoch": 1.3034190192414523,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022218446601941744,
      "loss": 0.5739,
      "step": 3227
    },
    {
      "epoch": 1.3038230392404424,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00022216019417475727,
      "loss": 0.6292,
      "step": 3228
    },
    {
      "epoch": 1.3042270592394323,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00022213592233009705,
      "loss": 0.6211,
      "step": 3229
    },
    {
      "epoch": 1.3046310792384224,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00022211165048543688,
      "loss": 0.5906,
      "step": 3230
    },
    {
      "epoch": 1.3050350992374122,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00022208737864077668,
      "loss": 0.6712,
      "step": 3231
    },
    {
      "epoch": 1.3054391192364023,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00022206310679611648,
      "loss": 0.6,
      "step": 3232
    },
    {
      "epoch": 1.3058431392353922,
      "grad_norm": 0.43359375,
      "learning_rate": 0.0002220388349514563,
      "loss": 0.5944,
      "step": 3233
    },
    {
      "epoch": 1.306247159234382,
      "grad_norm": 0.625,
      "learning_rate": 0.0002220145631067961,
      "loss": 0.6864,
      "step": 3234
    },
    {
      "epoch": 1.3066511792333722,
      "grad_norm": 0.44921875,
      "learning_rate": 0.0002219902912621359,
      "loss": 0.5949,
      "step": 3235
    },
    {
      "epoch": 1.307055199232362,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00022196601941747572,
      "loss": 0.5269,
      "step": 3236
    },
    {
      "epoch": 1.3074592192313519,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022194174757281552,
      "loss": 0.6366,
      "step": 3237
    },
    {
      "epoch": 1.307863239230342,
      "grad_norm": 0.46875,
      "learning_rate": 0.00022191747572815532,
      "loss": 0.6068,
      "step": 3238
    },
    {
      "epoch": 1.3082672592293318,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00022189320388349513,
      "loss": 0.5918,
      "step": 3239
    },
    {
      "epoch": 1.3086712792283217,
      "grad_norm": 0.5625,
      "learning_rate": 0.00022186893203883493,
      "loss": 0.5518,
      "step": 3240
    },
    {
      "epoch": 1.3090752992273118,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00022184466019417476,
      "loss": 0.5655,
      "step": 3241
    },
    {
      "epoch": 1.3094793192263017,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00022182038834951453,
      "loss": 0.623,
      "step": 3242
    },
    {
      "epoch": 1.3098833392252915,
      "grad_norm": 0.578125,
      "learning_rate": 0.00022179611650485434,
      "loss": 0.6746,
      "step": 3243
    },
    {
      "epoch": 1.3102873592242816,
      "grad_norm": 0.59375,
      "learning_rate": 0.00022177184466019417,
      "loss": 0.6491,
      "step": 3244
    },
    {
      "epoch": 1.3106913792232715,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00022174757281553397,
      "loss": 0.697,
      "step": 3245
    },
    {
      "epoch": 1.3110953992222614,
      "grad_norm": 0.625,
      "learning_rate": 0.00022172330097087374,
      "loss": 0.6655,
      "step": 3246
    },
    {
      "epoch": 1.3114994192212515,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00022169902912621357,
      "loss": 0.5603,
      "step": 3247
    },
    {
      "epoch": 1.3119034392202413,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022167475728155338,
      "loss": 0.6264,
      "step": 3248
    },
    {
      "epoch": 1.3123074592192314,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002216504854368932,
      "loss": 0.6287,
      "step": 3249
    },
    {
      "epoch": 1.3127114792182213,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00022162621359223298,
      "loss": 0.6727,
      "step": 3250
    },
    {
      "epoch": 1.3131154992172114,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00022160194174757278,
      "loss": 0.6166,
      "step": 3251
    },
    {
      "epoch": 1.3135195192162012,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002215776699029126,
      "loss": 0.5569,
      "step": 3252
    },
    {
      "epoch": 1.3139235392151911,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00022155339805825241,
      "loss": 0.6009,
      "step": 3253
    },
    {
      "epoch": 1.3143275592141812,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002215291262135922,
      "loss": 0.6322,
      "step": 3254
    },
    {
      "epoch": 1.314731579213171,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00022150485436893202,
      "loss": 0.6309,
      "step": 3255
    },
    {
      "epoch": 1.315135599212161,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00022148058252427182,
      "loss": 0.6345,
      "step": 3256
    },
    {
      "epoch": 1.315539619211151,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00022145631067961162,
      "loss": 0.6115,
      "step": 3257
    },
    {
      "epoch": 1.315943639210141,
      "grad_norm": 0.6875,
      "learning_rate": 0.00022143203883495145,
      "loss": 0.716,
      "step": 3258
    },
    {
      "epoch": 1.3163476592091308,
      "grad_norm": 27.625,
      "learning_rate": 0.00022140776699029123,
      "loss": 0.7329,
      "step": 3259
    },
    {
      "epoch": 1.3167516792081209,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00022138349514563106,
      "loss": 0.6029,
      "step": 3260
    },
    {
      "epoch": 1.3171556992071107,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00022135922330097086,
      "loss": 0.6529,
      "step": 3261
    },
    {
      "epoch": 1.3175597192061006,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00022133495145631066,
      "loss": 0.657,
      "step": 3262
    },
    {
      "epoch": 1.3179637392050907,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022131067961165046,
      "loss": 0.5948,
      "step": 3263
    },
    {
      "epoch": 1.3183677592040806,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00022128640776699027,
      "loss": 0.5842,
      "step": 3264
    },
    {
      "epoch": 1.3187717792030704,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00022126213592233007,
      "loss": 0.6079,
      "step": 3265
    },
    {
      "epoch": 1.3191757992020605,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002212378640776699,
      "loss": 0.6164,
      "step": 3266
    },
    {
      "epoch": 1.3195798192010504,
      "grad_norm": 0.59375,
      "learning_rate": 0.00022121359223300967,
      "loss": 0.6987,
      "step": 3267
    },
    {
      "epoch": 1.3199838392000405,
      "grad_norm": 0.41796875,
      "learning_rate": 0.0002211893203883495,
      "loss": 0.5473,
      "step": 3268
    },
    {
      "epoch": 1.3203878591990303,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002211650485436893,
      "loss": 0.6653,
      "step": 3269
    },
    {
      "epoch": 1.3207918791980204,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0002211407766990291,
      "loss": 0.6489,
      "step": 3270
    },
    {
      "epoch": 1.3211958991970103,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00022111650485436894,
      "loss": 0.5801,
      "step": 3271
    },
    {
      "epoch": 1.3215999191960002,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002210922330097087,
      "loss": 0.5383,
      "step": 3272
    },
    {
      "epoch": 1.3220039391949903,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00022106796116504852,
      "loss": 0.6694,
      "step": 3273
    },
    {
      "epoch": 1.3224079591939801,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00022104368932038834,
      "loss": 0.7713,
      "step": 3274
    },
    {
      "epoch": 1.32281197919297,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00022101941747572812,
      "loss": 0.6518,
      "step": 3275
    },
    {
      "epoch": 1.32321599919196,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00022099514563106792,
      "loss": 0.6441,
      "step": 3276
    },
    {
      "epoch": 1.32362001919095,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00022097087378640775,
      "loss": 0.59,
      "step": 3277
    },
    {
      "epoch": 1.3240240391899398,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00022094660194174755,
      "loss": 0.6417,
      "step": 3278
    },
    {
      "epoch": 1.32442805918893,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00022092233009708738,
      "loss": 0.6957,
      "step": 3279
    },
    {
      "epoch": 1.3248320791879198,
      "grad_norm": 0.6875,
      "learning_rate": 0.00022089805825242716,
      "loss": 0.66,
      "step": 3280
    },
    {
      "epoch": 1.3252360991869097,
      "grad_norm": 0.703125,
      "learning_rate": 0.00022087378640776696,
      "loss": 0.6415,
      "step": 3281
    },
    {
      "epoch": 1.3256401191858997,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002208495145631068,
      "loss": 0.6608,
      "step": 3282
    },
    {
      "epoch": 1.3260441391848896,
      "grad_norm": 0.625,
      "learning_rate": 0.0002208252427184466,
      "loss": 0.7031,
      "step": 3283
    },
    {
      "epoch": 1.3264481591838795,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00022080097087378637,
      "loss": 0.6195,
      "step": 3284
    },
    {
      "epoch": 1.3268521791828696,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002207766990291262,
      "loss": 0.5775,
      "step": 3285
    },
    {
      "epoch": 1.3272561991818594,
      "grad_norm": 0.52734375,
      "learning_rate": 0.000220752427184466,
      "loss": 0.6764,
      "step": 3286
    },
    {
      "epoch": 1.3276602191808495,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002207281553398058,
      "loss": 0.5957,
      "step": 3287
    },
    {
      "epoch": 1.3280642391798394,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002207038834951456,
      "loss": 0.6212,
      "step": 3288
    },
    {
      "epoch": 1.3284682591788293,
      "grad_norm": 48.25,
      "learning_rate": 0.0002206796116504854,
      "loss": 1.4445,
      "step": 3289
    },
    {
      "epoch": 1.3288722791778194,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00022065533980582524,
      "loss": 0.6644,
      "step": 3290
    },
    {
      "epoch": 1.3292762991768092,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022063106796116504,
      "loss": 0.6775,
      "step": 3291
    },
    {
      "epoch": 1.3296803191757993,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00022060679611650481,
      "loss": 0.7477,
      "step": 3292
    },
    {
      "epoch": 1.3300843391747892,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00022058252427184464,
      "loss": 0.6357,
      "step": 3293
    },
    {
      "epoch": 1.330488359173779,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00022055825242718445,
      "loss": 0.6383,
      "step": 3294
    },
    {
      "epoch": 1.3308923791727691,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00022053398058252425,
      "loss": 0.5942,
      "step": 3295
    },
    {
      "epoch": 1.331296399171759,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00022050970873786408,
      "loss": 0.68,
      "step": 3296
    },
    {
      "epoch": 1.3317004191707489,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00022048543689320385,
      "loss": 0.6227,
      "step": 3297
    },
    {
      "epoch": 1.332104439169739,
      "grad_norm": 0.578125,
      "learning_rate": 0.00022046116504854368,
      "loss": 0.6535,
      "step": 3298
    },
    {
      "epoch": 1.3325084591687288,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00022043689320388348,
      "loss": 0.6057,
      "step": 3299
    },
    {
      "epoch": 1.3329124791677187,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002204126213592233,
      "loss": 0.6233,
      "step": 3300
    },
    {
      "epoch": 1.3333164991667088,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002203883495145631,
      "loss": 0.6635,
      "step": 3301
    },
    {
      "epoch": 1.3337205191656987,
      "grad_norm": 0.671875,
      "learning_rate": 0.0002203640776699029,
      "loss": 0.6545,
      "step": 3302
    },
    {
      "epoch": 1.3341245391646885,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002203398058252427,
      "loss": 0.6351,
      "step": 3303
    },
    {
      "epoch": 1.3345285591636786,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00022031553398058252,
      "loss": 0.5497,
      "step": 3304
    },
    {
      "epoch": 1.3349325791626685,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002202912621359223,
      "loss": 0.6456,
      "step": 3305
    },
    {
      "epoch": 1.3353365991616584,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002202669902912621,
      "loss": 0.6802,
      "step": 3306
    },
    {
      "epoch": 1.3357406191606485,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00022024271844660193,
      "loss": 0.7348,
      "step": 3307
    },
    {
      "epoch": 1.3361446391596383,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00022021844660194173,
      "loss": 0.6553,
      "step": 3308
    },
    {
      "epoch": 1.3365486591586284,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00022019417475728156,
      "loss": 0.693,
      "step": 3309
    },
    {
      "epoch": 1.3369526791576183,
      "grad_norm": 0.578125,
      "learning_rate": 0.00022016990291262134,
      "loss": 0.6318,
      "step": 3310
    },
    {
      "epoch": 1.3373566991566084,
      "grad_norm": 0.42578125,
      "learning_rate": 0.00022014563106796114,
      "loss": 0.5965,
      "step": 3311
    },
    {
      "epoch": 1.3377607191555982,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00022012135922330097,
      "loss": 0.6854,
      "step": 3312
    },
    {
      "epoch": 1.3381647391545881,
      "grad_norm": 0.53125,
      "learning_rate": 0.00022009708737864074,
      "loss": 0.6571,
      "step": 3313
    },
    {
      "epoch": 1.3385687591535782,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00022007281553398055,
      "loss": 0.7127,
      "step": 3314
    },
    {
      "epoch": 1.338972779152568,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00022004854368932038,
      "loss": 0.6646,
      "step": 3315
    },
    {
      "epoch": 1.339376799151558,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00022002427184466018,
      "loss": 0.6238,
      "step": 3316
    },
    {
      "epoch": 1.339780819150548,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021999999999999995,
      "loss": 0.6446,
      "step": 3317
    },
    {
      "epoch": 1.340184839149538,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00021997572815533978,
      "loss": 0.6581,
      "step": 3318
    },
    {
      "epoch": 1.3405888591485278,
      "grad_norm": 4.875,
      "learning_rate": 0.00021995145631067959,
      "loss": 1.0206,
      "step": 3319
    },
    {
      "epoch": 1.3409928791475179,
      "grad_norm": 0.671875,
      "learning_rate": 0.00021992718446601941,
      "loss": 0.7179,
      "step": 3320
    },
    {
      "epoch": 1.3413968991465077,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00021990291262135922,
      "loss": 0.6647,
      "step": 3321
    },
    {
      "epoch": 1.3418009191454976,
      "grad_norm": 0.64453125,
      "learning_rate": 0.000219878640776699,
      "loss": 0.6259,
      "step": 3322
    },
    {
      "epoch": 1.3422049391444877,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00021985436893203882,
      "loss": 0.6533,
      "step": 3323
    },
    {
      "epoch": 1.3426089591434776,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00021983009708737862,
      "loss": 0.7159,
      "step": 3324
    },
    {
      "epoch": 1.3430129791424674,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00021980582524271843,
      "loss": 0.6158,
      "step": 3325
    },
    {
      "epoch": 1.3434169991414575,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021978155339805823,
      "loss": 0.6724,
      "step": 3326
    },
    {
      "epoch": 1.3438210191404474,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00021975728155339803,
      "loss": 0.6262,
      "step": 3327
    },
    {
      "epoch": 1.3442250391394375,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00021973300970873783,
      "loss": 0.6274,
      "step": 3328
    },
    {
      "epoch": 1.3446290591384273,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00021970873786407766,
      "loss": 0.617,
      "step": 3329
    },
    {
      "epoch": 1.3450330791374174,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00021968446601941744,
      "loss": 0.617,
      "step": 3330
    },
    {
      "epoch": 1.3454370991364073,
      "grad_norm": 0.578125,
      "learning_rate": 0.00021966019417475727,
      "loss": 0.6756,
      "step": 3331
    },
    {
      "epoch": 1.3458411191353972,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021963592233009707,
      "loss": 0.6644,
      "step": 3332
    },
    {
      "epoch": 1.3462451391343873,
      "grad_norm": 0.515625,
      "learning_rate": 0.00021961165048543687,
      "loss": 0.6016,
      "step": 3333
    },
    {
      "epoch": 1.3466491591333771,
      "grad_norm": 0.5,
      "learning_rate": 0.0002195873786407767,
      "loss": 0.6186,
      "step": 3334
    },
    {
      "epoch": 1.347053179132367,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00021956310679611648,
      "loss": 0.6838,
      "step": 3335
    },
    {
      "epoch": 1.347457199131357,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00021953883495145628,
      "loss": 0.7366,
      "step": 3336
    },
    {
      "epoch": 1.347861219130347,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002195145631067961,
      "loss": 0.6455,
      "step": 3337
    },
    {
      "epoch": 1.3482652391293368,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002194902912621359,
      "loss": 0.6124,
      "step": 3338
    },
    {
      "epoch": 1.348669259128327,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002194660194174757,
      "loss": 0.6827,
      "step": 3339
    },
    {
      "epoch": 1.3490732791273168,
      "grad_norm": 0.59375,
      "learning_rate": 0.00021944174757281552,
      "loss": 0.734,
      "step": 3340
    },
    {
      "epoch": 1.3494772991263067,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00021941747572815532,
      "loss": 0.7011,
      "step": 3341
    },
    {
      "epoch": 1.3498813191252967,
      "grad_norm": 0.59375,
      "learning_rate": 0.00021939320388349515,
      "loss": 0.6627,
      "step": 3342
    },
    {
      "epoch": 1.3502853391242866,
      "grad_norm": 0.53125,
      "learning_rate": 0.00021936893203883492,
      "loss": 0.6585,
      "step": 3343
    },
    {
      "epoch": 1.3506893591232765,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00021934466019417473,
      "loss": 0.644,
      "step": 3344
    },
    {
      "epoch": 1.3510933791222666,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00021932038834951455,
      "loss": 0.5599,
      "step": 3345
    },
    {
      "epoch": 1.3514973991212564,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00021929611650485436,
      "loss": 0.6526,
      "step": 3346
    },
    {
      "epoch": 1.3519014191202465,
      "grad_norm": 0.609375,
      "learning_rate": 0.00021927184466019413,
      "loss": 0.6558,
      "step": 3347
    },
    {
      "epoch": 1.3523054391192364,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00021924757281553396,
      "loss": 0.7175,
      "step": 3348
    },
    {
      "epoch": 1.3527094591182265,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00021922330097087376,
      "loss": 0.7383,
      "step": 3349
    },
    {
      "epoch": 1.3531134791172164,
      "grad_norm": 0.478515625,
      "learning_rate": 0.0002191990291262136,
      "loss": 0.5558,
      "step": 3350
    },
    {
      "epoch": 1.3535174991162062,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00021917475728155337,
      "loss": 0.5651,
      "step": 3351
    },
    {
      "epoch": 1.3539215191151963,
      "grad_norm": 0.546875,
      "learning_rate": 0.00021915048543689317,
      "loss": 0.5862,
      "step": 3352
    },
    {
      "epoch": 1.3543255391141862,
      "grad_norm": 0.55859375,
      "learning_rate": 0.000219126213592233,
      "loss": 0.6788,
      "step": 3353
    },
    {
      "epoch": 1.354729559113176,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002191019417475728,
      "loss": 0.6116,
      "step": 3354
    },
    {
      "epoch": 1.3551335791121661,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00021907766990291258,
      "loss": 0.6822,
      "step": 3355
    },
    {
      "epoch": 1.355537599111156,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002190533980582524,
      "loss": 0.7149,
      "step": 3356
    },
    {
      "epoch": 1.3559416191101459,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002190291262135922,
      "loss": 0.612,
      "step": 3357
    },
    {
      "epoch": 1.356345639109136,
      "grad_norm": 0.4453125,
      "learning_rate": 0.000219004854368932,
      "loss": 0.5689,
      "step": 3358
    },
    {
      "epoch": 1.3567496591081258,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00021898058252427184,
      "loss": 0.659,
      "step": 3359
    },
    {
      "epoch": 1.3571536791071157,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00021895631067961162,
      "loss": 0.5643,
      "step": 3360
    },
    {
      "epoch": 1.3575576991061058,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00021893203883495145,
      "loss": 0.5815,
      "step": 3361
    },
    {
      "epoch": 1.3579617191050957,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00021890776699029125,
      "loss": 0.563,
      "step": 3362
    },
    {
      "epoch": 1.3583657391040855,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00021888349514563105,
      "loss": 0.5777,
      "step": 3363
    },
    {
      "epoch": 1.3587697591030756,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00021885922330097085,
      "loss": 0.6447,
      "step": 3364
    },
    {
      "epoch": 1.3591737791020655,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021883495145631066,
      "loss": 0.6807,
      "step": 3365
    },
    {
      "epoch": 1.3595777991010556,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00021881067961165046,
      "loss": 0.7098,
      "step": 3366
    },
    {
      "epoch": 1.3599818191000455,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002187864077669903,
      "loss": 0.6288,
      "step": 3367
    },
    {
      "epoch": 1.3603858390990353,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00021876213592233006,
      "loss": 0.5953,
      "step": 3368
    },
    {
      "epoch": 1.3607898590980254,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002187378640776699,
      "loss": 0.6296,
      "step": 3369
    },
    {
      "epoch": 1.3611938790970153,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002187135922330097,
      "loss": 0.6725,
      "step": 3370
    },
    {
      "epoch": 1.3615978990960054,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002186893203883495,
      "loss": 0.6248,
      "step": 3371
    },
    {
      "epoch": 1.3620019190949952,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021866504854368933,
      "loss": 0.6538,
      "step": 3372
    },
    {
      "epoch": 1.362405939093985,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002186407766990291,
      "loss": 0.6246,
      "step": 3373
    },
    {
      "epoch": 1.3628099590929752,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002186165048543689,
      "loss": 0.5743,
      "step": 3374
    },
    {
      "epoch": 1.363213979091965,
      "grad_norm": 0.439453125,
      "learning_rate": 0.00021859223300970873,
      "loss": 0.587,
      "step": 3375
    },
    {
      "epoch": 1.363617999090955,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002185679611650485,
      "loss": 0.6488,
      "step": 3376
    },
    {
      "epoch": 1.364022019089945,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0002185436893203883,
      "loss": 0.7139,
      "step": 3377
    },
    {
      "epoch": 1.364426039088935,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00021851941747572814,
      "loss": 0.6657,
      "step": 3378
    },
    {
      "epoch": 1.3648300590879248,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00021849514563106794,
      "loss": 0.5943,
      "step": 3379
    },
    {
      "epoch": 1.3652340790869149,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00021847087378640777,
      "loss": 0.6827,
      "step": 3380
    },
    {
      "epoch": 1.3656380990859047,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00021844660194174755,
      "loss": 0.6146,
      "step": 3381
    },
    {
      "epoch": 1.3660421190848946,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00021842233009708735,
      "loss": 0.6962,
      "step": 3382
    },
    {
      "epoch": 1.3664461390838847,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00021839805825242718,
      "loss": 0.6343,
      "step": 3383
    },
    {
      "epoch": 1.3668501590828745,
      "grad_norm": 0.53125,
      "learning_rate": 0.00021837378640776698,
      "loss": 0.6747,
      "step": 3384
    },
    {
      "epoch": 1.3672541790818644,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021834951456310676,
      "loss": 0.6608,
      "step": 3385
    },
    {
      "epoch": 1.3676581990808545,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021832524271844659,
      "loss": 0.5884,
      "step": 3386
    },
    {
      "epoch": 1.3680622190798444,
      "grad_norm": 0.65625,
      "learning_rate": 0.0002183009708737864,
      "loss": 0.6625,
      "step": 3387
    },
    {
      "epoch": 1.3684662390788345,
      "grad_norm": 0.46875,
      "learning_rate": 0.0002182766990291262,
      "loss": 0.5706,
      "step": 3388
    },
    {
      "epoch": 1.3688702590778243,
      "grad_norm": 0.48046875,
      "learning_rate": 0.000218252427184466,
      "loss": 0.6568,
      "step": 3389
    },
    {
      "epoch": 1.3692742790768144,
      "grad_norm": 0.478515625,
      "learning_rate": 0.0002182281553398058,
      "loss": 0.6199,
      "step": 3390
    },
    {
      "epoch": 1.3696782990758043,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00021820388349514562,
      "loss": 0.6724,
      "step": 3391
    },
    {
      "epoch": 1.3700823190747942,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00021817961165048543,
      "loss": 0.6733,
      "step": 3392
    },
    {
      "epoch": 1.3704863390737843,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002181553398058252,
      "loss": 0.6903,
      "step": 3393
    },
    {
      "epoch": 1.3708903590727741,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00021813106796116503,
      "loss": 0.5983,
      "step": 3394
    },
    {
      "epoch": 1.371294379071764,
      "grad_norm": 0.671875,
      "learning_rate": 0.00021810679611650483,
      "loss": 0.7094,
      "step": 3395
    },
    {
      "epoch": 1.371698399070754,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00021808252427184464,
      "loss": 0.6468,
      "step": 3396
    },
    {
      "epoch": 1.372102419069744,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00021805825242718447,
      "loss": 0.6449,
      "step": 3397
    },
    {
      "epoch": 1.3725064390687338,
      "grad_norm": 0.431640625,
      "learning_rate": 0.00021803398058252424,
      "loss": 0.5523,
      "step": 3398
    },
    {
      "epoch": 1.372910459067724,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00021800970873786407,
      "loss": 0.6244,
      "step": 3399
    },
    {
      "epoch": 1.3733144790667138,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00021798543689320387,
      "loss": 0.6793,
      "step": 3400
    },
    {
      "epoch": 1.3737184990657036,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00021796116504854368,
      "loss": 0.7095,
      "step": 3401
    },
    {
      "epoch": 1.3741225190646937,
      "grad_norm": 0.59375,
      "learning_rate": 0.00021793689320388348,
      "loss": 0.674,
      "step": 3402
    },
    {
      "epoch": 1.3745265390636836,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00021791262135922328,
      "loss": 0.6394,
      "step": 3403
    },
    {
      "epoch": 1.3749305590626735,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00021788834951456308,
      "loss": 0.6218,
      "step": 3404
    },
    {
      "epoch": 1.3753345790616636,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002178640776699029,
      "loss": 0.67,
      "step": 3405
    },
    {
      "epoch": 1.3757385990606534,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002178398058252427,
      "loss": 0.6726,
      "step": 3406
    },
    {
      "epoch": 1.3761426190596435,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0002178155339805825,
      "loss": 0.6088,
      "step": 3407
    },
    {
      "epoch": 1.3765466390586334,
      "grad_norm": 0.515625,
      "learning_rate": 0.00021779126213592232,
      "loss": 0.5836,
      "step": 3408
    },
    {
      "epoch": 1.3769506590576235,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00021776699029126212,
      "loss": 0.6087,
      "step": 3409
    },
    {
      "epoch": 1.3773546790566134,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021774271844660195,
      "loss": 0.6519,
      "step": 3410
    },
    {
      "epoch": 1.3777586990556032,
      "grad_norm": 0.578125,
      "learning_rate": 0.00021771844660194173,
      "loss": 0.7228,
      "step": 3411
    },
    {
      "epoch": 1.3781627190545933,
      "grad_norm": 0.515625,
      "learning_rate": 0.00021769417475728153,
      "loss": 0.5863,
      "step": 3412
    },
    {
      "epoch": 1.3785667390535832,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00021766990291262136,
      "loss": 0.6195,
      "step": 3413
    },
    {
      "epoch": 1.378970759052573,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00021764563106796113,
      "loss": 0.6058,
      "step": 3414
    },
    {
      "epoch": 1.3793747790515631,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00021762135922330094,
      "loss": 0.6674,
      "step": 3415
    },
    {
      "epoch": 1.379778799050553,
      "grad_norm": 0.515625,
      "learning_rate": 0.00021759708737864076,
      "loss": 0.6522,
      "step": 3416
    },
    {
      "epoch": 1.3801828190495429,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00021757281553398057,
      "loss": 0.5214,
      "step": 3417
    },
    {
      "epoch": 1.380586839048533,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00021754854368932034,
      "loss": 0.6038,
      "step": 3418
    },
    {
      "epoch": 1.3809908590475228,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00021752427184466017,
      "loss": 0.7276,
      "step": 3419
    },
    {
      "epoch": 1.3813948790465127,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00021749999999999997,
      "loss": 0.6114,
      "step": 3420
    },
    {
      "epoch": 1.3817988990455028,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002174757281553398,
      "loss": 0.667,
      "step": 3421
    },
    {
      "epoch": 1.3822029190444927,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002174514563106796,
      "loss": 0.6705,
      "step": 3422
    },
    {
      "epoch": 1.3826069390434825,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00021742718446601938,
      "loss": 0.5662,
      "step": 3423
    },
    {
      "epoch": 1.3830109590424726,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002174029126213592,
      "loss": 0.602,
      "step": 3424
    },
    {
      "epoch": 1.3834149790414625,
      "grad_norm": 0.5,
      "learning_rate": 0.000217378640776699,
      "loss": 0.6181,
      "step": 3425
    },
    {
      "epoch": 1.3838189990404526,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00021735436893203882,
      "loss": 0.5713,
      "step": 3426
    },
    {
      "epoch": 1.3842230190394424,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00021733009708737862,
      "loss": 0.7062,
      "step": 3427
    },
    {
      "epoch": 1.3846270390384325,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00021730582524271842,
      "loss": 0.6063,
      "step": 3428
    },
    {
      "epoch": 1.3850310590374224,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00021728155339805822,
      "loss": 0.6148,
      "step": 3429
    },
    {
      "epoch": 1.3854350790364123,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00021725728155339805,
      "loss": 0.6125,
      "step": 3430
    },
    {
      "epoch": 1.3858390990354024,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00021723300970873783,
      "loss": 0.6041,
      "step": 3431
    },
    {
      "epoch": 1.3862431190343922,
      "grad_norm": 0.5,
      "learning_rate": 0.00021720873786407766,
      "loss": 0.5922,
      "step": 3432
    },
    {
      "epoch": 1.386647139033382,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00021718446601941746,
      "loss": 0.5636,
      "step": 3433
    },
    {
      "epoch": 1.3870511590323722,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00021716019417475726,
      "loss": 0.6363,
      "step": 3434
    },
    {
      "epoch": 1.387455179031362,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0002171359223300971,
      "loss": 0.7088,
      "step": 3435
    },
    {
      "epoch": 1.387859199030352,
      "grad_norm": 0.578125,
      "learning_rate": 0.00021711165048543687,
      "loss": 0.6349,
      "step": 3436
    },
    {
      "epoch": 1.388263219029342,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00021708737864077667,
      "loss": 0.6471,
      "step": 3437
    },
    {
      "epoch": 1.388667239028332,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002170631067961165,
      "loss": 0.6839,
      "step": 3438
    },
    {
      "epoch": 1.3890712590273218,
      "grad_norm": 0.53125,
      "learning_rate": 0.00021703883495145627,
      "loss": 0.6577,
      "step": 3439
    },
    {
      "epoch": 1.3894752790263118,
      "grad_norm": 0.484375,
      "learning_rate": 0.0002170145631067961,
      "loss": 0.5965,
      "step": 3440
    },
    {
      "epoch": 1.3898792990253017,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002169902912621359,
      "loss": 0.6096,
      "step": 3441
    },
    {
      "epoch": 1.3902833190242916,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0002169660194174757,
      "loss": 0.7264,
      "step": 3442
    },
    {
      "epoch": 1.3906873390232817,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00021694174757281554,
      "loss": 0.6694,
      "step": 3443
    },
    {
      "epoch": 1.3910913590222715,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002169174757281553,
      "loss": 0.7045,
      "step": 3444
    },
    {
      "epoch": 1.3914953790212614,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00021689320388349511,
      "loss": 0.6181,
      "step": 3445
    },
    {
      "epoch": 1.3918993990202515,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00021686893203883494,
      "loss": 0.623,
      "step": 3446
    },
    {
      "epoch": 1.3923034190192414,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00021684466019417475,
      "loss": 0.6434,
      "step": 3447
    },
    {
      "epoch": 1.3927074390182315,
      "grad_norm": 0.65625,
      "learning_rate": 0.00021682038834951452,
      "loss": 0.7365,
      "step": 3448
    },
    {
      "epoch": 1.3931114590172213,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021679611650485435,
      "loss": 0.6771,
      "step": 3449
    },
    {
      "epoch": 1.3935154790162114,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00021677184466019415,
      "loss": 0.6924,
      "step": 3450
    },
    {
      "epoch": 1.3939194990152013,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00021674757281553398,
      "loss": 0.7827,
      "step": 3451
    },
    {
      "epoch": 1.3943235190141912,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00021672330097087376,
      "loss": 0.6151,
      "step": 3452
    },
    {
      "epoch": 1.3947275390131813,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00021669902912621356,
      "loss": 0.656,
      "step": 3453
    },
    {
      "epoch": 1.3951315590121711,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002166747572815534,
      "loss": 0.6115,
      "step": 3454
    },
    {
      "epoch": 1.395535579011161,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002166504854368932,
      "loss": 0.6358,
      "step": 3455
    },
    {
      "epoch": 1.395939599010151,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00021662621359223297,
      "loss": 0.6226,
      "step": 3456
    },
    {
      "epoch": 1.396343619009141,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002166019417475728,
      "loss": 0.6212,
      "step": 3457
    },
    {
      "epoch": 1.3967476390081308,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0002165776699029126,
      "loss": 0.6065,
      "step": 3458
    },
    {
      "epoch": 1.397151659007121,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002165533980582524,
      "loss": 0.6571,
      "step": 3459
    },
    {
      "epoch": 1.3975556790061108,
      "grad_norm": 0.640625,
      "learning_rate": 0.00021652912621359223,
      "loss": 0.7419,
      "step": 3460
    },
    {
      "epoch": 1.3979596990051006,
      "grad_norm": 0.56640625,
      "learning_rate": 0.000216504854368932,
      "loss": 0.6586,
      "step": 3461
    },
    {
      "epoch": 1.3983637190040907,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00021648058252427184,
      "loss": 0.6186,
      "step": 3462
    },
    {
      "epoch": 1.3987677390030806,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021645631067961164,
      "loss": 0.6749,
      "step": 3463
    },
    {
      "epoch": 1.3991717590020705,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021643203883495144,
      "loss": 0.7164,
      "step": 3464
    },
    {
      "epoch": 1.3995757790010606,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00021640776699029124,
      "loss": 0.6068,
      "step": 3465
    },
    {
      "epoch": 1.3999797990000504,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021638349514563104,
      "loss": 0.5822,
      "step": 3466
    },
    {
      "epoch": 1.4003838189990405,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00021635922330097085,
      "loss": 0.682,
      "step": 3467
    },
    {
      "epoch": 1.4007878389980304,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00021633495145631068,
      "loss": 0.635,
      "step": 3468
    },
    {
      "epoch": 1.4011918589970205,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00021631067961165045,
      "loss": 0.5917,
      "step": 3469
    },
    {
      "epoch": 1.4015958789960103,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00021628640776699028,
      "loss": 0.6072,
      "step": 3470
    },
    {
      "epoch": 1.4019998989950002,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00021626213592233008,
      "loss": 0.6802,
      "step": 3471
    },
    {
      "epoch": 1.4024039189939903,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00021623786407766989,
      "loss": 0.684,
      "step": 3472
    },
    {
      "epoch": 1.4028079389929802,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00021621359223300972,
      "loss": 0.6759,
      "step": 3473
    },
    {
      "epoch": 1.40321195899197,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002161893203883495,
      "loss": 0.616,
      "step": 3474
    },
    {
      "epoch": 1.4036159789909601,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002161650485436893,
      "loss": 0.6577,
      "step": 3475
    },
    {
      "epoch": 1.40401999898995,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021614077669902912,
      "loss": 0.6014,
      "step": 3476
    },
    {
      "epoch": 1.4044240189889399,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002161165048543689,
      "loss": 0.5679,
      "step": 3477
    },
    {
      "epoch": 1.40482803898793,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002160922330097087,
      "loss": 0.5954,
      "step": 3478
    },
    {
      "epoch": 1.4052320589869198,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00021606796116504853,
      "loss": 0.6689,
      "step": 3479
    },
    {
      "epoch": 1.4056360789859097,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00021604368932038833,
      "loss": 0.6929,
      "step": 3480
    },
    {
      "epoch": 1.4060400989848998,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00021601941747572816,
      "loss": 0.6079,
      "step": 3481
    },
    {
      "epoch": 1.4064441189838897,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00021599514563106794,
      "loss": 0.6067,
      "step": 3482
    },
    {
      "epoch": 1.4068481389828795,
      "grad_norm": 0.77734375,
      "learning_rate": 0.00021597087378640774,
      "loss": 0.5938,
      "step": 3483
    },
    {
      "epoch": 1.4072521589818696,
      "grad_norm": 0.578125,
      "learning_rate": 0.00021594660194174757,
      "loss": 0.6693,
      "step": 3484
    },
    {
      "epoch": 1.4076561789808595,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00021592233009708737,
      "loss": 0.6134,
      "step": 3485
    },
    {
      "epoch": 1.4080601989798496,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00021589805825242715,
      "loss": 0.6228,
      "step": 3486
    },
    {
      "epoch": 1.4084642189788394,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00021587378640776697,
      "loss": 0.5946,
      "step": 3487
    },
    {
      "epoch": 1.4088682389778295,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00021584951456310678,
      "loss": 0.6884,
      "step": 3488
    },
    {
      "epoch": 1.4092722589768194,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021582524271844658,
      "loss": 0.6733,
      "step": 3489
    },
    {
      "epoch": 1.4096762789758093,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021580097087378638,
      "loss": 0.7001,
      "step": 3490
    },
    {
      "epoch": 1.4100802989747994,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00021577669902912618,
      "loss": 0.5942,
      "step": 3491
    },
    {
      "epoch": 1.4104843189737892,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00021575242718446601,
      "loss": 0.6846,
      "step": 3492
    },
    {
      "epoch": 1.410888338972779,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00021572815533980582,
      "loss": 0.6188,
      "step": 3493
    },
    {
      "epoch": 1.4112923589717692,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002157038834951456,
      "loss": 0.6219,
      "step": 3494
    },
    {
      "epoch": 1.411696378970759,
      "grad_norm": 0.451171875,
      "learning_rate": 0.00021567961165048542,
      "loss": 0.5974,
      "step": 3495
    },
    {
      "epoch": 1.412100398969749,
      "grad_norm": 0.431640625,
      "learning_rate": 0.00021565533980582522,
      "loss": 0.5647,
      "step": 3496
    },
    {
      "epoch": 1.412504418968739,
      "grad_norm": 0.625,
      "learning_rate": 0.00021563106796116503,
      "loss": 0.7005,
      "step": 3497
    },
    {
      "epoch": 1.4129084389677289,
      "grad_norm": 0.53125,
      "learning_rate": 0.00021560679611650485,
      "loss": 0.6336,
      "step": 3498
    },
    {
      "epoch": 1.4133124589667188,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00021558252427184463,
      "loss": 0.6206,
      "step": 3499
    },
    {
      "epoch": 1.4137164789657088,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00021555825242718446,
      "loss": 0.7263,
      "step": 3500
    },
    {
      "epoch": 1.4141204989646987,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021553398058252426,
      "loss": 0.6881,
      "step": 3501
    },
    {
      "epoch": 1.4145245189636886,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00021550970873786406,
      "loss": 0.6101,
      "step": 3502
    },
    {
      "epoch": 1.4149285389626787,
      "grad_norm": 0.65625,
      "learning_rate": 0.00021548543689320387,
      "loss": 0.6201,
      "step": 3503
    },
    {
      "epoch": 1.4153325589616685,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00021546116504854367,
      "loss": 0.6837,
      "step": 3504
    },
    {
      "epoch": 1.4157365789606586,
      "grad_norm": 0.578125,
      "learning_rate": 0.00021543689320388347,
      "loss": 0.641,
      "step": 3505
    },
    {
      "epoch": 1.4161405989596485,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002154126213592233,
      "loss": 0.6873,
      "step": 3506
    },
    {
      "epoch": 1.4165446189586386,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021538834951456308,
      "loss": 0.6894,
      "step": 3507
    },
    {
      "epoch": 1.4169486389576285,
      "grad_norm": 0.44921875,
      "learning_rate": 0.00021536407766990288,
      "loss": 0.5937,
      "step": 3508
    },
    {
      "epoch": 1.4173526589566183,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002153398058252427,
      "loss": 0.6313,
      "step": 3509
    },
    {
      "epoch": 1.4177566789556084,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002153155339805825,
      "loss": 0.6077,
      "step": 3510
    },
    {
      "epoch": 1.4181606989545983,
      "grad_norm": 0.53125,
      "learning_rate": 0.00021529126213592234,
      "loss": 0.5768,
      "step": 3511
    },
    {
      "epoch": 1.4185647189535882,
      "grad_norm": 0.484375,
      "learning_rate": 0.00021526699029126211,
      "loss": 0.6107,
      "step": 3512
    },
    {
      "epoch": 1.4189687389525782,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00021524271844660192,
      "loss": 0.5473,
      "step": 3513
    },
    {
      "epoch": 1.4193727589515681,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00021521844660194175,
      "loss": 0.6554,
      "step": 3514
    },
    {
      "epoch": 1.419776778950558,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00021519417475728152,
      "loss": 0.6759,
      "step": 3515
    },
    {
      "epoch": 1.420180798949548,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00021516990291262132,
      "loss": 0.6594,
      "step": 3516
    },
    {
      "epoch": 1.420584818948538,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00021514563106796115,
      "loss": 0.5934,
      "step": 3517
    },
    {
      "epoch": 1.4209888389475278,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021512135922330096,
      "loss": 0.6523,
      "step": 3518
    },
    {
      "epoch": 1.421392858946518,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021509708737864073,
      "loss": 0.631,
      "step": 3519
    },
    {
      "epoch": 1.4217968789455078,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021507281553398056,
      "loss": 0.6641,
      "step": 3520
    },
    {
      "epoch": 1.4222008989444976,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00021504854368932036,
      "loss": 0.6439,
      "step": 3521
    },
    {
      "epoch": 1.4226049189434877,
      "grad_norm": 0.451171875,
      "learning_rate": 0.0002150242718446602,
      "loss": 0.5728,
      "step": 3522
    },
    {
      "epoch": 1.4230089389424776,
      "grad_norm": 0.470703125,
      "learning_rate": 0.000215,
      "loss": 0.5853,
      "step": 3523
    },
    {
      "epoch": 1.4234129589414675,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021497572815533977,
      "loss": 0.6223,
      "step": 3524
    },
    {
      "epoch": 1.4238169789404576,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002149514563106796,
      "loss": 0.6293,
      "step": 3525
    },
    {
      "epoch": 1.4242209989394474,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002149271844660194,
      "loss": 0.6437,
      "step": 3526
    },
    {
      "epoch": 1.4246250189384375,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0002149029126213592,
      "loss": 0.5898,
      "step": 3527
    },
    {
      "epoch": 1.4250290389374274,
      "grad_norm": 0.57421875,
      "learning_rate": 0.000214878640776699,
      "loss": 0.6223,
      "step": 3528
    },
    {
      "epoch": 1.4254330589364175,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002148543689320388,
      "loss": 0.653,
      "step": 3529
    },
    {
      "epoch": 1.4258370789354073,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002148300970873786,
      "loss": 0.6817,
      "step": 3530
    },
    {
      "epoch": 1.4262410989343972,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00021480582524271844,
      "loss": 0.6204,
      "step": 3531
    },
    {
      "epoch": 1.4266451189333873,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00021478155339805822,
      "loss": 0.6449,
      "step": 3532
    },
    {
      "epoch": 1.4270491389323772,
      "grad_norm": 0.53125,
      "learning_rate": 0.00021475728155339805,
      "loss": 0.6323,
      "step": 3533
    },
    {
      "epoch": 1.427453158931367,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00021473300970873785,
      "loss": 0.677,
      "step": 3534
    },
    {
      "epoch": 1.4278571789303571,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00021470873786407765,
      "loss": 0.6069,
      "step": 3535
    },
    {
      "epoch": 1.428261198929347,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00021468446601941748,
      "loss": 0.5775,
      "step": 3536
    },
    {
      "epoch": 1.4286652189283369,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00021466019417475725,
      "loss": 0.6532,
      "step": 3537
    },
    {
      "epoch": 1.429069238927327,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00021463592233009706,
      "loss": 0.7,
      "step": 3538
    },
    {
      "epoch": 1.4294732589263168,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00021461165048543689,
      "loss": 0.5789,
      "step": 3539
    },
    {
      "epoch": 1.4298772789253067,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00021458737864077666,
      "loss": 0.7028,
      "step": 3540
    },
    {
      "epoch": 1.4302812989242968,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002145631067961165,
      "loss": 0.5638,
      "step": 3541
    },
    {
      "epoch": 1.4306853189232867,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002145388349514563,
      "loss": 0.5948,
      "step": 3542
    },
    {
      "epoch": 1.4310893389222765,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002145145631067961,
      "loss": 0.7497,
      "step": 3543
    },
    {
      "epoch": 1.4314933589212666,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00021449029126213593,
      "loss": 0.6616,
      "step": 3544
    },
    {
      "epoch": 1.4318973789202565,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0002144660194174757,
      "loss": 0.6583,
      "step": 3545
    },
    {
      "epoch": 1.4323013989192466,
      "grad_norm": 0.421875,
      "learning_rate": 0.0002144417475728155,
      "loss": 0.5986,
      "step": 3546
    },
    {
      "epoch": 1.4327054189182364,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00021441747572815533,
      "loss": 0.6203,
      "step": 3547
    },
    {
      "epoch": 1.4331094389172265,
      "grad_norm": 0.609375,
      "learning_rate": 0.00021439320388349513,
      "loss": 0.7094,
      "step": 3548
    },
    {
      "epoch": 1.4335134589162164,
      "grad_norm": 0.443359375,
      "learning_rate": 0.0002143689320388349,
      "loss": 0.5438,
      "step": 3549
    },
    {
      "epoch": 1.4339174789152063,
      "grad_norm": 0.484375,
      "learning_rate": 0.00021434466019417474,
      "loss": 0.5448,
      "step": 3550
    },
    {
      "epoch": 1.4343214989141964,
      "grad_norm": 0.416015625,
      "learning_rate": 0.00021432038834951454,
      "loss": 0.5497,
      "step": 3551
    },
    {
      "epoch": 1.4347255189131862,
      "grad_norm": 0.6875,
      "learning_rate": 0.00021429611650485437,
      "loss": 0.673,
      "step": 3552
    },
    {
      "epoch": 1.435129538912176,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00021427184466019415,
      "loss": 0.6937,
      "step": 3553
    },
    {
      "epoch": 1.4355335589111662,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00021424757281553395,
      "loss": 0.6438,
      "step": 3554
    },
    {
      "epoch": 1.435937578910156,
      "grad_norm": 0.71484375,
      "learning_rate": 0.00021422330097087378,
      "loss": 0.7469,
      "step": 3555
    },
    {
      "epoch": 1.436341598909146,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00021419902912621358,
      "loss": 0.6753,
      "step": 3556
    },
    {
      "epoch": 1.436745618908136,
      "grad_norm": 0.578125,
      "learning_rate": 0.00021417475728155336,
      "loss": 0.639,
      "step": 3557
    },
    {
      "epoch": 1.4371496389071259,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00021415048543689319,
      "loss": 0.6419,
      "step": 3558
    },
    {
      "epoch": 1.4375536589061157,
      "grad_norm": 0.55859375,
      "learning_rate": 0.000214126213592233,
      "loss": 0.6015,
      "step": 3559
    },
    {
      "epoch": 1.4379576789051058,
      "grad_norm": 0.44921875,
      "learning_rate": 0.0002141019417475728,
      "loss": 0.5926,
      "step": 3560
    },
    {
      "epoch": 1.4383616989040957,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00021407766990291262,
      "loss": 0.6175,
      "step": 3561
    },
    {
      "epoch": 1.4387657189030856,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002140533980582524,
      "loss": 0.6548,
      "step": 3562
    },
    {
      "epoch": 1.4391697389020757,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00021402912621359222,
      "loss": 0.5398,
      "step": 3563
    },
    {
      "epoch": 1.4395737589010655,
      "grad_norm": 0.625,
      "learning_rate": 0.00021400485436893203,
      "loss": 0.653,
      "step": 3564
    },
    {
      "epoch": 1.4399777789000556,
      "grad_norm": 0.6875,
      "learning_rate": 0.00021398058252427183,
      "loss": 0.6867,
      "step": 3565
    },
    {
      "epoch": 1.4403817988990455,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021395631067961163,
      "loss": 0.5803,
      "step": 3566
    },
    {
      "epoch": 1.4407858188980356,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00021393203883495143,
      "loss": 0.6925,
      "step": 3567
    },
    {
      "epoch": 1.4411898388970255,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00021390776699029124,
      "loss": 0.7536,
      "step": 3568
    },
    {
      "epoch": 1.4415938588960153,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00021388349514563107,
      "loss": 0.5707,
      "step": 3569
    },
    {
      "epoch": 1.4419978788950054,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00021385922330097084,
      "loss": 0.5808,
      "step": 3570
    },
    {
      "epoch": 1.4424018988939953,
      "grad_norm": 0.578125,
      "learning_rate": 0.00021383495145631067,
      "loss": 0.6727,
      "step": 3571
    },
    {
      "epoch": 1.4428059188929852,
      "grad_norm": 0.515625,
      "learning_rate": 0.00021381067961165047,
      "loss": 0.6688,
      "step": 3572
    },
    {
      "epoch": 1.4432099388919752,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00021378640776699027,
      "loss": 0.5913,
      "step": 3573
    },
    {
      "epoch": 1.443613958890965,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002137621359223301,
      "loss": 0.7058,
      "step": 3574
    },
    {
      "epoch": 1.444017978889955,
      "grad_norm": 0.515625,
      "learning_rate": 0.00021373786407766988,
      "loss": 0.6198,
      "step": 3575
    },
    {
      "epoch": 1.444421998888945,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00021371359223300968,
      "loss": 0.6136,
      "step": 3576
    },
    {
      "epoch": 1.444826018887935,
      "grad_norm": 0.67578125,
      "learning_rate": 0.0002136893203883495,
      "loss": 0.6812,
      "step": 3577
    },
    {
      "epoch": 1.4452300388869248,
      "grad_norm": 0.416015625,
      "learning_rate": 0.00021366504854368929,
      "loss": 0.5516,
      "step": 3578
    },
    {
      "epoch": 1.445634058885915,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0002136407766990291,
      "loss": 0.714,
      "step": 3579
    },
    {
      "epoch": 1.4460380788849048,
      "grad_norm": 0.546875,
      "learning_rate": 0.00021361650485436892,
      "loss": 0.6447,
      "step": 3580
    },
    {
      "epoch": 1.4464420988838946,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00021359223300970872,
      "loss": 0.5749,
      "step": 3581
    },
    {
      "epoch": 1.4468461188828847,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00021356796116504855,
      "loss": 0.5793,
      "step": 3582
    },
    {
      "epoch": 1.4472501388818746,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00021354368932038832,
      "loss": 0.7881,
      "step": 3583
    },
    {
      "epoch": 1.4476541588808647,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00021351941747572813,
      "loss": 0.6559,
      "step": 3584
    },
    {
      "epoch": 1.4480581788798546,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00021349514563106796,
      "loss": 0.5933,
      "step": 3585
    },
    {
      "epoch": 1.4484621988788444,
      "grad_norm": 0.515625,
      "learning_rate": 0.00021347087378640776,
      "loss": 0.5659,
      "step": 3586
    },
    {
      "epoch": 1.4488662188778345,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00021344660194174753,
      "loss": 0.6305,
      "step": 3587
    },
    {
      "epoch": 1.4492702388768244,
      "grad_norm": 0.671875,
      "learning_rate": 0.00021342233009708736,
      "loss": 0.6904,
      "step": 3588
    },
    {
      "epoch": 1.4496742588758145,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00021339805825242717,
      "loss": 0.5739,
      "step": 3589
    },
    {
      "epoch": 1.4500782788748043,
      "grad_norm": 0.609375,
      "learning_rate": 0.00021337378640776697,
      "loss": 0.6526,
      "step": 3590
    },
    {
      "epoch": 1.4504822988737942,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00021334951456310677,
      "loss": 0.6428,
      "step": 3591
    },
    {
      "epoch": 1.4508863188727843,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00021332524271844657,
      "loss": 0.4988,
      "step": 3592
    },
    {
      "epoch": 1.4512903388717742,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002133009708737864,
      "loss": 0.6553,
      "step": 3593
    },
    {
      "epoch": 1.451694358870764,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002132766990291262,
      "loss": 0.6279,
      "step": 3594
    },
    {
      "epoch": 1.4520983788697541,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00021325242718446598,
      "loss": 0.553,
      "step": 3595
    },
    {
      "epoch": 1.452502398868744,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002132281553398058,
      "loss": 0.6859,
      "step": 3596
    },
    {
      "epoch": 1.4529064188677339,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002132038834951456,
      "loss": 0.5649,
      "step": 3597
    },
    {
      "epoch": 1.453310438866724,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00021317961165048541,
      "loss": 0.5719,
      "step": 3598
    },
    {
      "epoch": 1.4537144588657138,
      "grad_norm": 0.53125,
      "learning_rate": 0.00021315533980582524,
      "loss": 0.6826,
      "step": 3599
    },
    {
      "epoch": 1.4541184788647037,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00021313106796116502,
      "loss": 0.6109,
      "step": 3600
    },
    {
      "epoch": 1.4545224988636938,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00021310679611650485,
      "loss": 0.6704,
      "step": 3601
    },
    {
      "epoch": 1.4549265188626836,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00021308252427184465,
      "loss": 0.5779,
      "step": 3602
    },
    {
      "epoch": 1.4553305388616735,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021305825242718443,
      "loss": 0.6682,
      "step": 3603
    },
    {
      "epoch": 1.4557345588606636,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00021303398058252426,
      "loss": 0.6642,
      "step": 3604
    },
    {
      "epoch": 1.4561385788596535,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00021300970873786406,
      "loss": 0.6886,
      "step": 3605
    },
    {
      "epoch": 1.4565425988586436,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021298543689320386,
      "loss": 0.6188,
      "step": 3606
    },
    {
      "epoch": 1.4569466188576334,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002129611650485437,
      "loss": 0.6772,
      "step": 3607
    },
    {
      "epoch": 1.4573506388566235,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00021293689320388346,
      "loss": 0.6517,
      "step": 3608
    },
    {
      "epoch": 1.4577546588556134,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00021291262135922327,
      "loss": 0.6936,
      "step": 3609
    },
    {
      "epoch": 1.4581586788546033,
      "grad_norm": 0.455078125,
      "learning_rate": 0.0002128883495145631,
      "loss": 0.5793,
      "step": 3610
    },
    {
      "epoch": 1.4585626988535934,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0002128640776699029,
      "loss": 0.7104,
      "step": 3611
    },
    {
      "epoch": 1.4589667188525832,
      "grad_norm": 0.5,
      "learning_rate": 0.00021283980582524273,
      "loss": 0.6165,
      "step": 3612
    },
    {
      "epoch": 1.459370738851573,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002128155339805825,
      "loss": 0.6439,
      "step": 3613
    },
    {
      "epoch": 1.4597747588505632,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002127912621359223,
      "loss": 0.6538,
      "step": 3614
    },
    {
      "epoch": 1.460178778849553,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00021276699029126214,
      "loss": 0.6224,
      "step": 3615
    },
    {
      "epoch": 1.460582798848543,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002127427184466019,
      "loss": 0.5657,
      "step": 3616
    },
    {
      "epoch": 1.460986818847533,
      "grad_norm": 0.423828125,
      "learning_rate": 0.0002127184466019417,
      "loss": 0.5849,
      "step": 3617
    },
    {
      "epoch": 1.4613908388465229,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00021269417475728154,
      "loss": 0.6263,
      "step": 3618
    },
    {
      "epoch": 1.4617948588455127,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00021266990291262134,
      "loss": 0.6518,
      "step": 3619
    },
    {
      "epoch": 1.4621988788445028,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00021264563106796112,
      "loss": 0.6036,
      "step": 3620
    },
    {
      "epoch": 1.4626028988434927,
      "grad_norm": 0.6875,
      "learning_rate": 0.00021262135922330095,
      "loss": 0.7846,
      "step": 3621
    },
    {
      "epoch": 1.4630069188424826,
      "grad_norm": 0.76171875,
      "learning_rate": 0.00021259708737864075,
      "loss": 0.6917,
      "step": 3622
    },
    {
      "epoch": 1.4634109388414727,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00021257281553398058,
      "loss": 0.6263,
      "step": 3623
    },
    {
      "epoch": 1.4638149588404625,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00021254854368932038,
      "loss": 0.5791,
      "step": 3624
    },
    {
      "epoch": 1.4642189788394526,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00021252427184466016,
      "loss": 0.5416,
      "step": 3625
    },
    {
      "epoch": 1.4646229988384425,
      "grad_norm": 0.515625,
      "learning_rate": 0.0002125,
      "loss": 0.6606,
      "step": 3626
    },
    {
      "epoch": 1.4650270188374326,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0002124757281553398,
      "loss": 0.7093,
      "step": 3627
    },
    {
      "epoch": 1.4654310388364225,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002124514563106796,
      "loss": 0.6371,
      "step": 3628
    },
    {
      "epoch": 1.4658350588354123,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002124271844660194,
      "loss": 0.5894,
      "step": 3629
    },
    {
      "epoch": 1.4662390788344024,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002124029126213592,
      "loss": 0.6012,
      "step": 3630
    },
    {
      "epoch": 1.4666430988333923,
      "grad_norm": 0.453125,
      "learning_rate": 0.000212378640776699,
      "loss": 0.5589,
      "step": 3631
    },
    {
      "epoch": 1.4670471188323821,
      "grad_norm": 0.59375,
      "learning_rate": 0.00021235436893203883,
      "loss": 0.6295,
      "step": 3632
    },
    {
      "epoch": 1.4674511388313722,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002123300970873786,
      "loss": 0.6238,
      "step": 3633
    },
    {
      "epoch": 1.467855158830362,
      "grad_norm": 0.515625,
      "learning_rate": 0.00021230582524271843,
      "loss": 0.6208,
      "step": 3634
    },
    {
      "epoch": 1.468259178829352,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00021228155339805824,
      "loss": 0.6059,
      "step": 3635
    },
    {
      "epoch": 1.468663198828342,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00021225728155339804,
      "loss": 0.6787,
      "step": 3636
    },
    {
      "epoch": 1.469067218827332,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00021223300970873787,
      "loss": 0.6307,
      "step": 3637
    },
    {
      "epoch": 1.4694712388263218,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00021220873786407764,
      "loss": 0.658,
      "step": 3638
    },
    {
      "epoch": 1.469875258825312,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00021218446601941745,
      "loss": 0.5611,
      "step": 3639
    },
    {
      "epoch": 1.4702792788243018,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00021216019417475728,
      "loss": 0.5641,
      "step": 3640
    },
    {
      "epoch": 1.4706832988232916,
      "grad_norm": 0.59375,
      "learning_rate": 0.00021213592233009705,
      "loss": 0.6724,
      "step": 3641
    },
    {
      "epoch": 1.4710873188222817,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00021211165048543688,
      "loss": 0.6522,
      "step": 3642
    },
    {
      "epoch": 1.4714913388212716,
      "grad_norm": 0.546875,
      "learning_rate": 0.00021208737864077668,
      "loss": 0.6802,
      "step": 3643
    },
    {
      "epoch": 1.4718953588202617,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00021206310679611648,
      "loss": 0.6283,
      "step": 3644
    },
    {
      "epoch": 1.4722993788192515,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00021203883495145631,
      "loss": 0.6528,
      "step": 3645
    },
    {
      "epoch": 1.4727033988182416,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002120145631067961,
      "loss": 0.5767,
      "step": 3646
    },
    {
      "epoch": 1.4731074188172315,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002119902912621359,
      "loss": 0.6531,
      "step": 3647
    },
    {
      "epoch": 1.4735114388162214,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00021196601941747572,
      "loss": 0.684,
      "step": 3648
    },
    {
      "epoch": 1.4739154588152115,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00021194174757281552,
      "loss": 0.6947,
      "step": 3649
    },
    {
      "epoch": 1.4743194788142013,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002119174757281553,
      "loss": 0.626,
      "step": 3650
    },
    {
      "epoch": 1.4747234988131912,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00021189320388349513,
      "loss": 0.6168,
      "step": 3651
    },
    {
      "epoch": 1.4751275188121813,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00021186893203883493,
      "loss": 0.7477,
      "step": 3652
    },
    {
      "epoch": 1.4755315388111712,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00021184466019417476,
      "loss": 0.6507,
      "step": 3653
    },
    {
      "epoch": 1.475935558810161,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00021182038834951454,
      "loss": 0.6085,
      "step": 3654
    },
    {
      "epoch": 1.4763395788091511,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00021179611650485434,
      "loss": 0.6112,
      "step": 3655
    },
    {
      "epoch": 1.476743598808141,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00021177184466019417,
      "loss": 0.655,
      "step": 3656
    },
    {
      "epoch": 1.4771476188071309,
      "grad_norm": 0.671875,
      "learning_rate": 0.00021174757281553397,
      "loss": 0.6717,
      "step": 3657
    },
    {
      "epoch": 1.477551638806121,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021172330097087374,
      "loss": 0.5964,
      "step": 3658
    },
    {
      "epoch": 1.4779556588051108,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00021169902912621357,
      "loss": 0.5808,
      "step": 3659
    },
    {
      "epoch": 1.4783596788041007,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00021167475728155338,
      "loss": 0.5842,
      "step": 3660
    },
    {
      "epoch": 1.4787636988030908,
      "grad_norm": 0.625,
      "learning_rate": 0.00021165048543689318,
      "loss": 0.7122,
      "step": 3661
    },
    {
      "epoch": 1.4791677188020806,
      "grad_norm": 0.498046875,
      "learning_rate": 0.000211626213592233,
      "loss": 0.5631,
      "step": 3662
    },
    {
      "epoch": 1.4795717388010707,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00021160194174757278,
      "loss": 0.6216,
      "step": 3663
    },
    {
      "epoch": 1.4799757588000606,
      "grad_norm": 0.4453125,
      "learning_rate": 0.0002115776699029126,
      "loss": 0.5553,
      "step": 3664
    },
    {
      "epoch": 1.4803797787990505,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00021155339805825241,
      "loss": 0.718,
      "step": 3665
    },
    {
      "epoch": 1.4807837987980406,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00021152912621359222,
      "loss": 0.6176,
      "step": 3666
    },
    {
      "epoch": 1.4811878187970304,
      "grad_norm": 0.515625,
      "learning_rate": 0.00021150485436893202,
      "loss": 0.6233,
      "step": 3667
    },
    {
      "epoch": 1.4815918387960205,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00021148058252427182,
      "loss": 0.6342,
      "step": 3668
    },
    {
      "epoch": 1.4819958587950104,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021145631067961162,
      "loss": 0.7509,
      "step": 3669
    },
    {
      "epoch": 1.4823998787940003,
      "grad_norm": 0.439453125,
      "learning_rate": 0.00021143203883495145,
      "loss": 0.5741,
      "step": 3670
    },
    {
      "epoch": 1.4828038987929903,
      "grad_norm": 0.416015625,
      "learning_rate": 0.00021140776699029123,
      "loss": 0.6421,
      "step": 3671
    },
    {
      "epoch": 1.4832079187919802,
      "grad_norm": 0.53125,
      "learning_rate": 0.00021138349514563106,
      "loss": 0.6128,
      "step": 3672
    },
    {
      "epoch": 1.48361193879097,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00021135922330097086,
      "loss": 0.6015,
      "step": 3673
    },
    {
      "epoch": 1.4840159587899602,
      "grad_norm": 0.5625,
      "learning_rate": 0.00021133495145631066,
      "loss": 0.5858,
      "step": 3674
    },
    {
      "epoch": 1.48441997878895,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002113106796116505,
      "loss": 0.6693,
      "step": 3675
    },
    {
      "epoch": 1.48482399878794,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00021128640776699027,
      "loss": 0.695,
      "step": 3676
    },
    {
      "epoch": 1.48522801878693,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00021126213592233007,
      "loss": 0.5803,
      "step": 3677
    },
    {
      "epoch": 1.4856320387859199,
      "grad_norm": 0.4375,
      "learning_rate": 0.0002112378640776699,
      "loss": 0.5651,
      "step": 3678
    },
    {
      "epoch": 1.4860360587849097,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00021121359223300967,
      "loss": 0.5878,
      "step": 3679
    },
    {
      "epoch": 1.4864400787838998,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00021118932038834948,
      "loss": 0.5718,
      "step": 3680
    },
    {
      "epoch": 1.4868440987828897,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0002111650485436893,
      "loss": 0.6372,
      "step": 3681
    },
    {
      "epoch": 1.4872481187818796,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002111407766990291,
      "loss": 0.7082,
      "step": 3682
    },
    {
      "epoch": 1.4876521387808697,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00021111650485436894,
      "loss": 0.5532,
      "step": 3683
    },
    {
      "epoch": 1.4880561587798595,
      "grad_norm": 0.640625,
      "learning_rate": 0.00021109223300970871,
      "loss": 0.6145,
      "step": 3684
    },
    {
      "epoch": 1.4884601787788496,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00021106796116504852,
      "loss": 0.6445,
      "step": 3685
    },
    {
      "epoch": 1.4888641987778395,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00021104368932038835,
      "loss": 0.6516,
      "step": 3686
    },
    {
      "epoch": 1.4892682187768296,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00021101941747572815,
      "loss": 0.6069,
      "step": 3687
    },
    {
      "epoch": 1.4896722387758194,
      "grad_norm": 0.578125,
      "learning_rate": 0.00021099514563106792,
      "loss": 0.6603,
      "step": 3688
    },
    {
      "epoch": 1.4900762587748093,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00021097087378640775,
      "loss": 0.653,
      "step": 3689
    },
    {
      "epoch": 1.4904802787737994,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00021094660194174755,
      "loss": 0.6048,
      "step": 3690
    },
    {
      "epoch": 1.4908842987727893,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00021092233009708736,
      "loss": 0.617,
      "step": 3691
    },
    {
      "epoch": 1.4912883187717791,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021089805825242716,
      "loss": 0.6622,
      "step": 3692
    },
    {
      "epoch": 1.4916923387707692,
      "grad_norm": 0.5,
      "learning_rate": 0.00021087378640776696,
      "loss": 0.591,
      "step": 3693
    },
    {
      "epoch": 1.492096358769759,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0002108495145631068,
      "loss": 0.71,
      "step": 3694
    },
    {
      "epoch": 1.492500378768749,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002108252427184466,
      "loss": 0.6588,
      "step": 3695
    },
    {
      "epoch": 1.492904398767739,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00021080097087378637,
      "loss": 0.6478,
      "step": 3696
    },
    {
      "epoch": 1.493308418766729,
      "grad_norm": 0.44140625,
      "learning_rate": 0.0002107766990291262,
      "loss": 0.6151,
      "step": 3697
    },
    {
      "epoch": 1.4937124387657188,
      "grad_norm": 0.6015625,
      "learning_rate": 0.000210752427184466,
      "loss": 0.665,
      "step": 3698
    },
    {
      "epoch": 1.4941164587647089,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0002107281553398058,
      "loss": 0.8003,
      "step": 3699
    },
    {
      "epoch": 1.4945204787636988,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00021070388349514563,
      "loss": 0.5502,
      "step": 3700
    },
    {
      "epoch": 1.4949244987626886,
      "grad_norm": 0.71875,
      "learning_rate": 0.0002106796116504854,
      "loss": 0.6302,
      "step": 3701
    },
    {
      "epoch": 1.4953285187616787,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00021065533980582524,
      "loss": 0.5996,
      "step": 3702
    },
    {
      "epoch": 1.4957325387606686,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00021063106796116504,
      "loss": 0.6498,
      "step": 3703
    },
    {
      "epoch": 1.4961365587596587,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00021060679611650481,
      "loss": 0.6314,
      "step": 3704
    },
    {
      "epoch": 1.4965405787586485,
      "grad_norm": 0.44140625,
      "learning_rate": 0.00021058252427184464,
      "loss": 0.5495,
      "step": 3705
    },
    {
      "epoch": 1.4969445987576386,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021055825242718445,
      "loss": 0.7145,
      "step": 3706
    },
    {
      "epoch": 1.4973486187566285,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00021053398058252425,
      "loss": 0.5876,
      "step": 3707
    },
    {
      "epoch": 1.4977526387556184,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00021050970873786408,
      "loss": 0.6417,
      "step": 3708
    },
    {
      "epoch": 1.4981566587546085,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00021048543689320385,
      "loss": 0.6894,
      "step": 3709
    },
    {
      "epoch": 1.4985606787535983,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00021046116504854366,
      "loss": 0.5751,
      "step": 3710
    },
    {
      "epoch": 1.4989646987525882,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00021043689320388349,
      "loss": 0.6748,
      "step": 3711
    },
    {
      "epoch": 1.4993687187515783,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0002104126213592233,
      "loss": 0.6441,
      "step": 3712
    },
    {
      "epoch": 1.4997727387505682,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00021038834951456312,
      "loss": 0.6742,
      "step": 3713
    },
    {
      "epoch": 1.500176758749558,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002103640776699029,
      "loss": 0.7085,
      "step": 3714
    },
    {
      "epoch": 1.5005807787485481,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002103398058252427,
      "loss": 0.5641,
      "step": 3715
    },
    {
      "epoch": 1.500984798747538,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00021031553398058252,
      "loss": 0.5913,
      "step": 3716
    },
    {
      "epoch": 1.5013888187465279,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002102912621359223,
      "loss": 0.6452,
      "step": 3717
    },
    {
      "epoch": 1.501792838745518,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002102669902912621,
      "loss": 0.6829,
      "step": 3718
    },
    {
      "epoch": 1.5021968587445078,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00021024271844660193,
      "loss": 0.6481,
      "step": 3719
    },
    {
      "epoch": 1.5026008787434977,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00021021844660194173,
      "loss": 0.6092,
      "step": 3720
    },
    {
      "epoch": 1.5030048987424878,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002101941747572815,
      "loss": 0.641,
      "step": 3721
    },
    {
      "epoch": 1.5034089187414779,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00021016990291262134,
      "loss": 0.58,
      "step": 3722
    },
    {
      "epoch": 1.5038129387404675,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00021014563106796114,
      "loss": 0.6462,
      "step": 3723
    },
    {
      "epoch": 1.5042169587394576,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00021012135922330097,
      "loss": 0.6807,
      "step": 3724
    },
    {
      "epoch": 1.5046209787384477,
      "grad_norm": 0.546875,
      "learning_rate": 0.00021009708737864077,
      "loss": 0.6564,
      "step": 3725
    },
    {
      "epoch": 1.5050249987374376,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00021007281553398055,
      "loss": 0.7164,
      "step": 3726
    },
    {
      "epoch": 1.5054290187364274,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00021004854368932038,
      "loss": 0.6551,
      "step": 3727
    },
    {
      "epoch": 1.5058330387354175,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00021002427184466018,
      "loss": 0.6212,
      "step": 3728
    },
    {
      "epoch": 1.5062370587344074,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020999999999999998,
      "loss": 0.6548,
      "step": 3729
    },
    {
      "epoch": 1.5066410787333973,
      "grad_norm": 0.59375,
      "learning_rate": 0.00020997572815533978,
      "loss": 0.6163,
      "step": 3730
    },
    {
      "epoch": 1.5070450987323873,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00020995145631067959,
      "loss": 0.7089,
      "step": 3731
    },
    {
      "epoch": 1.5074491187313772,
      "grad_norm": 0.46484375,
      "learning_rate": 0.0002099271844660194,
      "loss": 0.614,
      "step": 3732
    },
    {
      "epoch": 1.507853138730367,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00020990291262135922,
      "loss": 0.6541,
      "step": 3733
    },
    {
      "epoch": 1.5082571587293572,
      "grad_norm": 0.515625,
      "learning_rate": 0.000209878640776699,
      "loss": 0.616,
      "step": 3734
    },
    {
      "epoch": 1.508661178728347,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00020985436893203882,
      "loss": 0.5896,
      "step": 3735
    },
    {
      "epoch": 1.509065198727337,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020983009708737863,
      "loss": 0.6135,
      "step": 3736
    },
    {
      "epoch": 1.509469218726327,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020980582524271843,
      "loss": 0.6765,
      "step": 3737
    },
    {
      "epoch": 1.5098732387253169,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020978155339805826,
      "loss": 0.6006,
      "step": 3738
    },
    {
      "epoch": 1.5102772587243067,
      "grad_norm": 0.5,
      "learning_rate": 0.00020975728155339803,
      "loss": 0.6095,
      "step": 3739
    },
    {
      "epoch": 1.5106812787232968,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00020973300970873783,
      "loss": 0.5988,
      "step": 3740
    },
    {
      "epoch": 1.511085298722287,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020970873786407766,
      "loss": 0.6968,
      "step": 3741
    },
    {
      "epoch": 1.5114893187212766,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00020968446601941744,
      "loss": 0.628,
      "step": 3742
    },
    {
      "epoch": 1.5118933387202667,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00020966019417475727,
      "loss": 0.7147,
      "step": 3743
    },
    {
      "epoch": 1.5122973587192567,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00020963592233009707,
      "loss": 0.6873,
      "step": 3744
    },
    {
      "epoch": 1.5127013787182464,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00020961165048543687,
      "loss": 0.656,
      "step": 3745
    },
    {
      "epoch": 1.5131053987172365,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002095873786407767,
      "loss": 0.617,
      "step": 3746
    },
    {
      "epoch": 1.5135094187162266,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00020956310679611648,
      "loss": 0.6524,
      "step": 3747
    },
    {
      "epoch": 1.5139134387152164,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00020953883495145628,
      "loss": 0.5901,
      "step": 3748
    },
    {
      "epoch": 1.5143174587142063,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002095145631067961,
      "loss": 0.6639,
      "step": 3749
    },
    {
      "epoch": 1.5147214787131964,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0002094902912621359,
      "loss": 0.6903,
      "step": 3750
    },
    {
      "epoch": 1.5151254987121863,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002094660194174757,
      "loss": 0.705,
      "step": 3751
    },
    {
      "epoch": 1.5155295187111761,
      "grad_norm": 0.5625,
      "learning_rate": 0.00020944174757281552,
      "loss": 0.6682,
      "step": 3752
    },
    {
      "epoch": 1.5159335387101662,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00020941747572815532,
      "loss": 0.6563,
      "step": 3753
    },
    {
      "epoch": 1.516337558709156,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020939320388349515,
      "loss": 0.5872,
      "step": 3754
    },
    {
      "epoch": 1.516741578708146,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00020936893203883492,
      "loss": 0.685,
      "step": 3755
    },
    {
      "epoch": 1.517145598707136,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00020934466019417473,
      "loss": 0.6657,
      "step": 3756
    },
    {
      "epoch": 1.517549618706126,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00020932038834951456,
      "loss": 0.5917,
      "step": 3757
    },
    {
      "epoch": 1.5179536387051158,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00020929611650485436,
      "loss": 0.6301,
      "step": 3758
    },
    {
      "epoch": 1.5183576587041059,
      "grad_norm": 0.44921875,
      "learning_rate": 0.00020927184466019413,
      "loss": 0.5756,
      "step": 3759
    },
    {
      "epoch": 1.5187616787030958,
      "grad_norm": 0.427734375,
      "learning_rate": 0.00020924757281553396,
      "loss": 0.542,
      "step": 3760
    },
    {
      "epoch": 1.5191656987020856,
      "grad_norm": 0.4140625,
      "learning_rate": 0.00020922330097087376,
      "loss": 0.5964,
      "step": 3761
    },
    {
      "epoch": 1.5195697187010757,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00020919902912621357,
      "loss": 0.5952,
      "step": 3762
    },
    {
      "epoch": 1.5199737387000658,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002091747572815534,
      "loss": 0.6672,
      "step": 3763
    },
    {
      "epoch": 1.5203777586990554,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020915048543689317,
      "loss": 0.6473,
      "step": 3764
    },
    {
      "epoch": 1.5207817786980455,
      "grad_norm": 0.546875,
      "learning_rate": 0.000209126213592233,
      "loss": 0.6768,
      "step": 3765
    },
    {
      "epoch": 1.5211857986970356,
      "grad_norm": 0.6484375,
      "learning_rate": 0.0002091019417475728,
      "loss": 0.664,
      "step": 3766
    },
    {
      "epoch": 1.5215898186960255,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00020907766990291258,
      "loss": 0.6748,
      "step": 3767
    },
    {
      "epoch": 1.5219938386950154,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002090533980582524,
      "loss": 0.6665,
      "step": 3768
    },
    {
      "epoch": 1.5223978586940055,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002090291262135922,
      "loss": 0.6028,
      "step": 3769
    },
    {
      "epoch": 1.5228018786929953,
      "grad_norm": 0.53515625,
      "learning_rate": 0.000209004854368932,
      "loss": 0.673,
      "step": 3770
    },
    {
      "epoch": 1.5232058986919852,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00020898058252427184,
      "loss": 0.6203,
      "step": 3771
    },
    {
      "epoch": 1.5236099186909753,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00020895631067961162,
      "loss": 0.6295,
      "step": 3772
    },
    {
      "epoch": 1.5240139386899652,
      "grad_norm": 0.484375,
      "learning_rate": 0.00020893203883495145,
      "loss": 0.6095,
      "step": 3773
    },
    {
      "epoch": 1.524417958688955,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020890776699029125,
      "loss": 0.5955,
      "step": 3774
    },
    {
      "epoch": 1.5248219786879451,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00020888349514563105,
      "loss": 0.5402,
      "step": 3775
    },
    {
      "epoch": 1.525225998686935,
      "grad_norm": 0.578125,
      "learning_rate": 0.00020885922330097088,
      "loss": 0.7139,
      "step": 3776
    },
    {
      "epoch": 1.5256300186859248,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00020883495145631066,
      "loss": 0.6138,
      "step": 3777
    },
    {
      "epoch": 1.526034038684915,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00020881067961165046,
      "loss": 0.6105,
      "step": 3778
    },
    {
      "epoch": 1.5264380586839048,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002087864077669903,
      "loss": 0.6215,
      "step": 3779
    },
    {
      "epoch": 1.5268420786828947,
      "grad_norm": 0.44921875,
      "learning_rate": 0.00020876213592233006,
      "loss": 0.5421,
      "step": 3780
    },
    {
      "epoch": 1.5272460986818848,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00020873786407766987,
      "loss": 0.6858,
      "step": 3781
    },
    {
      "epoch": 1.5276501186808749,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002087135922330097,
      "loss": 0.5097,
      "step": 3782
    },
    {
      "epoch": 1.5280541386798645,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002086893203883495,
      "loss": 0.6151,
      "step": 3783
    },
    {
      "epoch": 1.5284581586788546,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00020866504854368933,
      "loss": 0.5974,
      "step": 3784
    },
    {
      "epoch": 1.5288621786778447,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002086407766990291,
      "loss": 0.5967,
      "step": 3785
    },
    {
      "epoch": 1.5292661986768346,
      "grad_norm": 0.640625,
      "learning_rate": 0.0002086165048543689,
      "loss": 0.6958,
      "step": 3786
    },
    {
      "epoch": 1.5296702186758244,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00020859223300970873,
      "loss": 0.6156,
      "step": 3787
    },
    {
      "epoch": 1.5300742386748145,
      "grad_norm": 0.451171875,
      "learning_rate": 0.00020856796116504854,
      "loss": 0.5857,
      "step": 3788
    },
    {
      "epoch": 1.5304782586738044,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0002085436893203883,
      "loss": 0.7216,
      "step": 3789
    },
    {
      "epoch": 1.5308822786727942,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00020851941747572814,
      "loss": 0.6874,
      "step": 3790
    },
    {
      "epoch": 1.5312862986717843,
      "grad_norm": 0.546875,
      "learning_rate": 0.00020849514563106794,
      "loss": 0.6131,
      "step": 3791
    },
    {
      "epoch": 1.5316903186707742,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00020847087378640775,
      "loss": 0.6565,
      "step": 3792
    },
    {
      "epoch": 1.532094338669764,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00020844660194174755,
      "loss": 0.7061,
      "step": 3793
    },
    {
      "epoch": 1.5324983586687542,
      "grad_norm": 0.703125,
      "learning_rate": 0.00020842233009708735,
      "loss": 0.682,
      "step": 3794
    },
    {
      "epoch": 1.532902378667744,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020839805825242718,
      "loss": 0.6224,
      "step": 3795
    },
    {
      "epoch": 1.533306398666734,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00020837378640776698,
      "loss": 0.6348,
      "step": 3796
    },
    {
      "epoch": 1.533710418665724,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00020834951456310676,
      "loss": 0.6318,
      "step": 3797
    },
    {
      "epoch": 1.5341144386647139,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002083252427184466,
      "loss": 0.6876,
      "step": 3798
    },
    {
      "epoch": 1.5345184586637037,
      "grad_norm": 0.734375,
      "learning_rate": 0.0002083009708737864,
      "loss": 0.6783,
      "step": 3799
    },
    {
      "epoch": 1.5349224786626938,
      "grad_norm": 0.46484375,
      "learning_rate": 0.0002082766990291262,
      "loss": 0.5544,
      "step": 3800
    },
    {
      "epoch": 1.535326498661684,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00020825242718446602,
      "loss": 0.616,
      "step": 3801
    },
    {
      "epoch": 1.5357305186606736,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002082281553398058,
      "loss": 0.7117,
      "step": 3802
    },
    {
      "epoch": 1.5361345386596637,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00020820388349514563,
      "loss": 0.6313,
      "step": 3803
    },
    {
      "epoch": 1.5365385586586537,
      "grad_norm": 0.53125,
      "learning_rate": 0.00020817961165048543,
      "loss": 0.6413,
      "step": 3804
    },
    {
      "epoch": 1.5369425786576434,
      "grad_norm": 0.5,
      "learning_rate": 0.0002081553398058252,
      "loss": 0.6089,
      "step": 3805
    },
    {
      "epoch": 1.5373465986566335,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00020813106796116503,
      "loss": 0.6374,
      "step": 3806
    },
    {
      "epoch": 1.5377506186556236,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020810679611650484,
      "loss": 0.6443,
      "step": 3807
    },
    {
      "epoch": 1.5381546386546134,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00020808252427184464,
      "loss": 0.638,
      "step": 3808
    },
    {
      "epoch": 1.5385586586536033,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00020805825242718447,
      "loss": 0.6072,
      "step": 3809
    },
    {
      "epoch": 1.5389626786525934,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00020803398058252424,
      "loss": 0.6473,
      "step": 3810
    },
    {
      "epoch": 1.5393666986515833,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00020800970873786404,
      "loss": 0.702,
      "step": 3811
    },
    {
      "epoch": 1.5397707186505731,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00020798543689320387,
      "loss": 0.6719,
      "step": 3812
    },
    {
      "epoch": 1.5401747386495632,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00020796116504854368,
      "loss": 0.702,
      "step": 3813
    },
    {
      "epoch": 1.540578758648553,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002079368932038835,
      "loss": 0.7115,
      "step": 3814
    },
    {
      "epoch": 1.540982778647543,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00020791262135922328,
      "loss": 0.6068,
      "step": 3815
    },
    {
      "epoch": 1.541386798646533,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00020788834951456308,
      "loss": 0.6245,
      "step": 3816
    },
    {
      "epoch": 1.541790818645523,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0002078640776699029,
      "loss": 0.643,
      "step": 3817
    },
    {
      "epoch": 1.5421948386445128,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002078398058252427,
      "loss": 0.6509,
      "step": 3818
    },
    {
      "epoch": 1.5425988586435029,
      "grad_norm": 0.4375,
      "learning_rate": 0.0002078155339805825,
      "loss": 0.59,
      "step": 3819
    },
    {
      "epoch": 1.543002878642493,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00020779126213592232,
      "loss": 0.5924,
      "step": 3820
    },
    {
      "epoch": 1.5434068986414826,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00020776699029126212,
      "loss": 0.6229,
      "step": 3821
    },
    {
      "epoch": 1.5438109186404727,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002077427184466019,
      "loss": 0.5962,
      "step": 3822
    },
    {
      "epoch": 1.5442149386394628,
      "grad_norm": 0.609375,
      "learning_rate": 0.00020771844660194173,
      "loss": 0.6826,
      "step": 3823
    },
    {
      "epoch": 1.5446189586384524,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020769417475728153,
      "loss": 0.6108,
      "step": 3824
    },
    {
      "epoch": 1.5450229786374425,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00020766990291262136,
      "loss": 0.6445,
      "step": 3825
    },
    {
      "epoch": 1.5454269986364326,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00020764563106796116,
      "loss": 0.6325,
      "step": 3826
    },
    {
      "epoch": 1.5458310186354225,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00020762135922330094,
      "loss": 0.733,
      "step": 3827
    },
    {
      "epoch": 1.5462350386344124,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00020759708737864077,
      "loss": 0.653,
      "step": 3828
    },
    {
      "epoch": 1.5466390586334025,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00020757281553398057,
      "loss": 0.6301,
      "step": 3829
    },
    {
      "epoch": 1.5470430786323923,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00020754854368932037,
      "loss": 0.6981,
      "step": 3830
    },
    {
      "epoch": 1.5474470986313822,
      "grad_norm": 0.546875,
      "learning_rate": 0.00020752427184466017,
      "loss": 0.6348,
      "step": 3831
    },
    {
      "epoch": 1.5478511186303723,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00020749999999999998,
      "loss": 0.6017,
      "step": 3832
    },
    {
      "epoch": 1.5482551386293621,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020747572815533978,
      "loss": 0.649,
      "step": 3833
    },
    {
      "epoch": 1.548659158628352,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002074514563106796,
      "loss": 0.6309,
      "step": 3834
    },
    {
      "epoch": 1.549063178627342,
      "grad_norm": 0.546875,
      "learning_rate": 0.00020742718446601938,
      "loss": 0.6274,
      "step": 3835
    },
    {
      "epoch": 1.549467198626332,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002074029126213592,
      "loss": 0.7542,
      "step": 3836
    },
    {
      "epoch": 1.5498712186253218,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00020737864077669901,
      "loss": 0.5998,
      "step": 3837
    },
    {
      "epoch": 1.550275238624312,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00020735436893203882,
      "loss": 0.6202,
      "step": 3838
    },
    {
      "epoch": 1.5506792586233018,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020733009708737865,
      "loss": 0.5876,
      "step": 3839
    },
    {
      "epoch": 1.5510832786222917,
      "grad_norm": 0.59375,
      "learning_rate": 0.00020730582524271842,
      "loss": 0.6564,
      "step": 3840
    },
    {
      "epoch": 1.5514872986212818,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00020728155339805822,
      "loss": 0.5879,
      "step": 3841
    },
    {
      "epoch": 1.5518913186202719,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00020725728155339805,
      "loss": 0.575,
      "step": 3842
    },
    {
      "epoch": 1.5522953386192615,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00020723300970873783,
      "loss": 0.6303,
      "step": 3843
    },
    {
      "epoch": 1.5526993586182516,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00020720873786407766,
      "loss": 0.6327,
      "step": 3844
    },
    {
      "epoch": 1.5531033786172417,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00020718446601941746,
      "loss": 0.5899,
      "step": 3845
    },
    {
      "epoch": 1.5535073986162315,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020716019417475726,
      "loss": 0.6741,
      "step": 3846
    },
    {
      "epoch": 1.5539114186152214,
      "grad_norm": 0.41796875,
      "learning_rate": 0.0002071359223300971,
      "loss": 0.5437,
      "step": 3847
    },
    {
      "epoch": 1.5543154386142115,
      "grad_norm": 0.5625,
      "learning_rate": 0.00020711165048543687,
      "loss": 0.7247,
      "step": 3848
    },
    {
      "epoch": 1.5547194586132014,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00020708737864077667,
      "loss": 0.7495,
      "step": 3849
    },
    {
      "epoch": 1.5551234786121912,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002070631067961165,
      "loss": 0.6384,
      "step": 3850
    },
    {
      "epoch": 1.5555274986111813,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0002070388349514563,
      "loss": 0.6434,
      "step": 3851
    },
    {
      "epoch": 1.5559315186101712,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00020701456310679608,
      "loss": 0.6262,
      "step": 3852
    },
    {
      "epoch": 1.556335538609161,
      "grad_norm": 0.48046875,
      "learning_rate": 0.0002069902912621359,
      "loss": 0.5567,
      "step": 3853
    },
    {
      "epoch": 1.5567395586081512,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002069660194174757,
      "loss": 0.5745,
      "step": 3854
    },
    {
      "epoch": 1.557143578607141,
      "grad_norm": 0.46875,
      "learning_rate": 0.00020694174757281554,
      "loss": 0.5862,
      "step": 3855
    },
    {
      "epoch": 1.557547598606131,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002069174757281553,
      "loss": 0.6507,
      "step": 3856
    },
    {
      "epoch": 1.557951618605121,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00020689320388349511,
      "loss": 0.6274,
      "step": 3857
    },
    {
      "epoch": 1.5583556386041109,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00020686893203883494,
      "loss": 0.6829,
      "step": 3858
    },
    {
      "epoch": 1.5587596586031007,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00020684466019417475,
      "loss": 0.6406,
      "step": 3859
    },
    {
      "epoch": 1.5591636786020908,
      "grad_norm": 0.46875,
      "learning_rate": 0.00020682038834951452,
      "loss": 0.6709,
      "step": 3860
    },
    {
      "epoch": 1.559567698601081,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020679611650485435,
      "loss": 0.6642,
      "step": 3861
    },
    {
      "epoch": 1.5599717186000706,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00020677184466019415,
      "loss": 0.5737,
      "step": 3862
    },
    {
      "epoch": 1.5603757385990606,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00020674757281553396,
      "loss": 0.6487,
      "step": 3863
    },
    {
      "epoch": 1.5607797585980507,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00020672330097087379,
      "loss": 0.6651,
      "step": 3864
    },
    {
      "epoch": 1.5611837785970406,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00020669902912621356,
      "loss": 0.7214,
      "step": 3865
    },
    {
      "epoch": 1.5615877985960305,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0002066747572815534,
      "loss": 0.6017,
      "step": 3866
    },
    {
      "epoch": 1.5619918185950206,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0002066504854368932,
      "loss": 0.656,
      "step": 3867
    },
    {
      "epoch": 1.5623958385940104,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00020662621359223297,
      "loss": 0.6941,
      "step": 3868
    },
    {
      "epoch": 1.5627998585930003,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002066019417475728,
      "loss": 0.5422,
      "step": 3869
    },
    {
      "epoch": 1.5632038785919904,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0002065776699029126,
      "loss": 0.6178,
      "step": 3870
    },
    {
      "epoch": 1.5636078985909803,
      "grad_norm": 0.4453125,
      "learning_rate": 0.0002065533980582524,
      "loss": 0.5417,
      "step": 3871
    },
    {
      "epoch": 1.5640119185899701,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00020652912621359223,
      "loss": 0.6172,
      "step": 3872
    },
    {
      "epoch": 1.5644159385889602,
      "grad_norm": 0.55078125,
      "learning_rate": 0.000206504854368932,
      "loss": 0.6617,
      "step": 3873
    },
    {
      "epoch": 1.56481995858795,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00020648058252427184,
      "loss": 0.5799,
      "step": 3874
    },
    {
      "epoch": 1.56522397858694,
      "grad_norm": 0.42578125,
      "learning_rate": 0.00020645631067961164,
      "loss": 0.5487,
      "step": 3875
    },
    {
      "epoch": 1.56562799858593,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020643203883495144,
      "loss": 0.6072,
      "step": 3876
    },
    {
      "epoch": 1.56603201858492,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00020640776699029127,
      "loss": 0.5958,
      "step": 3877
    },
    {
      "epoch": 1.5664360385839098,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00020638349514563105,
      "loss": 0.5823,
      "step": 3878
    },
    {
      "epoch": 1.5668400585828999,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00020635922330097085,
      "loss": 0.73,
      "step": 3879
    },
    {
      "epoch": 1.56724407858189,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00020633495145631068,
      "loss": 0.6375,
      "step": 3880
    },
    {
      "epoch": 1.5676480985808796,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00020631067961165045,
      "loss": 0.6462,
      "step": 3881
    },
    {
      "epoch": 1.5680521185798697,
      "grad_norm": 0.484375,
      "learning_rate": 0.00020628640776699025,
      "loss": 0.5897,
      "step": 3882
    },
    {
      "epoch": 1.5684561385788598,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020626213592233008,
      "loss": 0.6266,
      "step": 3883
    },
    {
      "epoch": 1.5688601585778494,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002062378640776699,
      "loss": 0.6802,
      "step": 3884
    },
    {
      "epoch": 1.5692641785768395,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00020621359223300972,
      "loss": 0.6325,
      "step": 3885
    },
    {
      "epoch": 1.5696681985758296,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002061893203883495,
      "loss": 0.608,
      "step": 3886
    },
    {
      "epoch": 1.5700722185748195,
      "grad_norm": 0.462890625,
      "learning_rate": 0.0002061650485436893,
      "loss": 0.551,
      "step": 3887
    },
    {
      "epoch": 1.5704762385738094,
      "grad_norm": 0.53125,
      "learning_rate": 0.00020614077669902912,
      "loss": 0.6607,
      "step": 3888
    },
    {
      "epoch": 1.5708802585727994,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020611650485436893,
      "loss": 0.6326,
      "step": 3889
    },
    {
      "epoch": 1.5712842785717893,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0002060922330097087,
      "loss": 0.628,
      "step": 3890
    },
    {
      "epoch": 1.5716882985707792,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00020606796116504853,
      "loss": 0.6616,
      "step": 3891
    },
    {
      "epoch": 1.5720923185697693,
      "grad_norm": 0.640625,
      "learning_rate": 0.00020604368932038833,
      "loss": 0.6626,
      "step": 3892
    },
    {
      "epoch": 1.5724963385687591,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00020601941747572813,
      "loss": 0.6102,
      "step": 3893
    },
    {
      "epoch": 1.572900358567749,
      "grad_norm": 0.546875,
      "learning_rate": 0.00020599514563106794,
      "loss": 0.6207,
      "step": 3894
    },
    {
      "epoch": 1.573304378566739,
      "grad_norm": 1.625,
      "learning_rate": 0.00020597087378640774,
      "loss": 0.6629,
      "step": 3895
    },
    {
      "epoch": 1.573708398565729,
      "grad_norm": 0.431640625,
      "learning_rate": 0.00020594660194174757,
      "loss": 0.5822,
      "step": 3896
    },
    {
      "epoch": 1.5741124185647188,
      "grad_norm": 0.546875,
      "learning_rate": 0.00020592233009708737,
      "loss": 0.6405,
      "step": 3897
    },
    {
      "epoch": 1.574516438563709,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00020589805825242715,
      "loss": 0.7791,
      "step": 3898
    },
    {
      "epoch": 1.5749204585626988,
      "grad_norm": 0.609375,
      "learning_rate": 0.00020587378640776698,
      "loss": 0.6579,
      "step": 3899
    },
    {
      "epoch": 1.5753244785616887,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00020584951456310678,
      "loss": 0.6349,
      "step": 3900
    },
    {
      "epoch": 1.5757284985606788,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020582524271844658,
      "loss": 0.6549,
      "step": 3901
    },
    {
      "epoch": 1.5761325185596688,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0002058009708737864,
      "loss": 0.5699,
      "step": 3902
    },
    {
      "epoch": 1.5765365385586585,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00020577669902912619,
      "loss": 0.6531,
      "step": 3903
    },
    {
      "epoch": 1.5769405585576486,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00020575242718446601,
      "loss": 0.5481,
      "step": 3904
    },
    {
      "epoch": 1.5773445785566387,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00020572815533980582,
      "loss": 0.6711,
      "step": 3905
    },
    {
      "epoch": 1.5777485985556285,
      "grad_norm": 0.4296875,
      "learning_rate": 0.0002057038834951456,
      "loss": 0.5609,
      "step": 3906
    },
    {
      "epoch": 1.5781526185546184,
      "grad_norm": 0.453125,
      "learning_rate": 0.00020567961165048542,
      "loss": 0.6308,
      "step": 3907
    },
    {
      "epoch": 1.5785566385536085,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00020565533980582522,
      "loss": 0.6594,
      "step": 3908
    },
    {
      "epoch": 1.5789606585525984,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00020563106796116503,
      "loss": 0.6345,
      "step": 3909
    },
    {
      "epoch": 1.5793646785515882,
      "grad_norm": 0.46875,
      "learning_rate": 0.00020560679611650486,
      "loss": 0.5328,
      "step": 3910
    },
    {
      "epoch": 1.5797686985505783,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020558252427184463,
      "loss": 0.6168,
      "step": 3911
    },
    {
      "epoch": 1.5801727185495682,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00020555825242718443,
      "loss": 0.7012,
      "step": 3912
    },
    {
      "epoch": 1.580576738548558,
      "grad_norm": 0.53125,
      "learning_rate": 0.00020553398058252426,
      "loss": 0.6596,
      "step": 3913
    },
    {
      "epoch": 1.5809807585475482,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00020550970873786407,
      "loss": 0.5898,
      "step": 3914
    },
    {
      "epoch": 1.581384778546538,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020548543689320387,
      "loss": 0.618,
      "step": 3915
    },
    {
      "epoch": 1.581788798545528,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00020546116504854367,
      "loss": 0.5734,
      "step": 3916
    },
    {
      "epoch": 1.582192818544518,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00020543689320388347,
      "loss": 0.625,
      "step": 3917
    },
    {
      "epoch": 1.5825968385435079,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002054126213592233,
      "loss": 0.6831,
      "step": 3918
    },
    {
      "epoch": 1.5830008585424977,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00020538834951456308,
      "loss": 0.6203,
      "step": 3919
    },
    {
      "epoch": 1.5834048785414878,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00020536407766990288,
      "loss": 0.7314,
      "step": 3920
    },
    {
      "epoch": 1.583808898540478,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0002053398058252427,
      "loss": 0.7615,
      "step": 3921
    },
    {
      "epoch": 1.5842129185394676,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002053155339805825,
      "loss": 0.6298,
      "step": 3922
    },
    {
      "epoch": 1.5846169385384576,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00020529126213592229,
      "loss": 0.7101,
      "step": 3923
    },
    {
      "epoch": 1.5850209585374477,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020526699029126212,
      "loss": 0.6217,
      "step": 3924
    },
    {
      "epoch": 1.5854249785364376,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00020524271844660192,
      "loss": 0.6128,
      "step": 3925
    },
    {
      "epoch": 1.5858289985354275,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00020521844660194175,
      "loss": 0.6123,
      "step": 3926
    },
    {
      "epoch": 1.5862330185344176,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00020519417475728155,
      "loss": 0.6941,
      "step": 3927
    },
    {
      "epoch": 1.5866370385334074,
      "grad_norm": 0.53125,
      "learning_rate": 0.00020516990291262132,
      "loss": 0.6615,
      "step": 3928
    },
    {
      "epoch": 1.5870410585323973,
      "grad_norm": 0.609375,
      "learning_rate": 0.00020514563106796115,
      "loss": 0.5408,
      "step": 3929
    },
    {
      "epoch": 1.5874450785313874,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00020512135922330096,
      "loss": 0.5992,
      "step": 3930
    },
    {
      "epoch": 1.5878490985303773,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00020509708737864073,
      "loss": 0.641,
      "step": 3931
    },
    {
      "epoch": 1.5882531185293671,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00020507281553398056,
      "loss": 0.6089,
      "step": 3932
    },
    {
      "epoch": 1.5886571385283572,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00020504854368932036,
      "loss": 0.5982,
      "step": 3933
    },
    {
      "epoch": 1.589061158527347,
      "grad_norm": 0.609375,
      "learning_rate": 0.0002050242718446602,
      "loss": 0.5614,
      "step": 3934
    },
    {
      "epoch": 1.589465178526337,
      "grad_norm": 0.53515625,
      "learning_rate": 0.000205,
      "loss": 0.5966,
      "step": 3935
    },
    {
      "epoch": 1.589869198525327,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00020497572815533977,
      "loss": 0.6466,
      "step": 3936
    },
    {
      "epoch": 1.590273218524317,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002049514563106796,
      "loss": 0.5528,
      "step": 3937
    },
    {
      "epoch": 1.5906772385233068,
      "grad_norm": 0.66796875,
      "learning_rate": 0.0002049271844660194,
      "loss": 0.6176,
      "step": 3938
    },
    {
      "epoch": 1.5910812585222969,
      "grad_norm": 0.46875,
      "learning_rate": 0.0002049029126213592,
      "loss": 0.6234,
      "step": 3939
    },
    {
      "epoch": 1.591485278521287,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020487864077669903,
      "loss": 0.5724,
      "step": 3940
    },
    {
      "epoch": 1.5918892985202766,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002048543689320388,
      "loss": 0.6259,
      "step": 3941
    },
    {
      "epoch": 1.5922933185192667,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002048300970873786,
      "loss": 0.6012,
      "step": 3942
    },
    {
      "epoch": 1.5926973385182568,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00020480582524271844,
      "loss": 0.6161,
      "step": 3943
    },
    {
      "epoch": 1.5931013585172467,
      "grad_norm": 0.53125,
      "learning_rate": 0.00020478155339805822,
      "loss": 0.7152,
      "step": 3944
    },
    {
      "epoch": 1.5935053785162365,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00020475728155339805,
      "loss": 0.6497,
      "step": 3945
    },
    {
      "epoch": 1.5939093985152266,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00020473300970873785,
      "loss": 0.7031,
      "step": 3946
    },
    {
      "epoch": 1.5943134185142165,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020470873786407765,
      "loss": 0.6416,
      "step": 3947
    },
    {
      "epoch": 1.5947174385132064,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00020468446601941748,
      "loss": 0.7407,
      "step": 3948
    },
    {
      "epoch": 1.5951214585121964,
      "grad_norm": 0.435546875,
      "learning_rate": 0.00020466019417475726,
      "loss": 0.5733,
      "step": 3949
    },
    {
      "epoch": 1.5955254785111863,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00020463592233009706,
      "loss": 0.5785,
      "step": 3950
    },
    {
      "epoch": 1.5959294985101762,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002046116504854369,
      "loss": 0.6206,
      "step": 3951
    },
    {
      "epoch": 1.5963335185091663,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002045873786407767,
      "loss": 0.6514,
      "step": 3952
    },
    {
      "epoch": 1.5967375385081561,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00020456310679611646,
      "loss": 0.627,
      "step": 3953
    },
    {
      "epoch": 1.597141558507146,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0002045388349514563,
      "loss": 0.6799,
      "step": 3954
    },
    {
      "epoch": 1.597545578506136,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002045145631067961,
      "loss": 0.7212,
      "step": 3955
    },
    {
      "epoch": 1.597949598505126,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00020449029126213593,
      "loss": 0.5971,
      "step": 3956
    },
    {
      "epoch": 1.5983536185041158,
      "grad_norm": 0.578125,
      "learning_rate": 0.0002044660194174757,
      "loss": 0.636,
      "step": 3957
    },
    {
      "epoch": 1.598757638503106,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0002044417475728155,
      "loss": 0.6414,
      "step": 3958
    },
    {
      "epoch": 1.599161658502096,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00020441747572815533,
      "loss": 0.7094,
      "step": 3959
    },
    {
      "epoch": 1.5995656785010857,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00020439320388349514,
      "loss": 0.5592,
      "step": 3960
    },
    {
      "epoch": 1.5999696985000758,
      "grad_norm": 0.458984375,
      "learning_rate": 0.0002043689320388349,
      "loss": 0.6281,
      "step": 3961
    },
    {
      "epoch": 1.6003737184990658,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00020434466019417474,
      "loss": 0.615,
      "step": 3962
    },
    {
      "epoch": 1.6007777384980555,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00020432038834951454,
      "loss": 0.6312,
      "step": 3963
    },
    {
      "epoch": 1.6011817584970456,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020429611650485434,
      "loss": 0.5859,
      "step": 3964
    },
    {
      "epoch": 1.6015857784960357,
      "grad_norm": 0.53125,
      "learning_rate": 0.00020427184466019417,
      "loss": 0.6863,
      "step": 3965
    },
    {
      "epoch": 1.6019897984950255,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020424757281553395,
      "loss": 0.5888,
      "step": 3966
    },
    {
      "epoch": 1.6023938184940154,
      "grad_norm": 0.68359375,
      "learning_rate": 0.00020422330097087378,
      "loss": 0.6954,
      "step": 3967
    },
    {
      "epoch": 1.6027978384930055,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00020419902912621358,
      "loss": 0.6476,
      "step": 3968
    },
    {
      "epoch": 1.6032018584919954,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00020417475728155336,
      "loss": 0.6177,
      "step": 3969
    },
    {
      "epoch": 1.6036058784909852,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00020415048543689319,
      "loss": 0.6916,
      "step": 3970
    },
    {
      "epoch": 1.6040098984899753,
      "grad_norm": 0.5078125,
      "learning_rate": 0.000204126213592233,
      "loss": 0.6222,
      "step": 3971
    },
    {
      "epoch": 1.6044139184889652,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002041019417475728,
      "loss": 0.6812,
      "step": 3972
    },
    {
      "epoch": 1.604817938487955,
      "grad_norm": 0.609375,
      "learning_rate": 0.00020407766990291262,
      "loss": 0.6573,
      "step": 3973
    },
    {
      "epoch": 1.6052219584869452,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002040533980582524,
      "loss": 0.5921,
      "step": 3974
    },
    {
      "epoch": 1.605625978485935,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00020402912621359222,
      "loss": 0.628,
      "step": 3975
    },
    {
      "epoch": 1.606029998484925,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00020400485436893203,
      "loss": 0.5745,
      "step": 3976
    },
    {
      "epoch": 1.606434018483915,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020398058252427183,
      "loss": 0.6059,
      "step": 3977
    },
    {
      "epoch": 1.6068380384829049,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00020395631067961163,
      "loss": 0.6315,
      "step": 3978
    },
    {
      "epoch": 1.6072420584818947,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00020393203883495143,
      "loss": 0.662,
      "step": 3979
    },
    {
      "epoch": 1.6076460784808848,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00020390776699029124,
      "loss": 0.639,
      "step": 3980
    },
    {
      "epoch": 1.608050098479875,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00020388349514563107,
      "loss": 0.5664,
      "step": 3981
    },
    {
      "epoch": 1.6084541184788645,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020385922330097084,
      "loss": 0.6522,
      "step": 3982
    },
    {
      "epoch": 1.6088581384778546,
      "grad_norm": 0.5625,
      "learning_rate": 0.00020383495145631064,
      "loss": 0.6732,
      "step": 3983
    },
    {
      "epoch": 1.6092621584768447,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020381067961165047,
      "loss": 0.4989,
      "step": 3984
    },
    {
      "epoch": 1.6096661784758346,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020378640776699028,
      "loss": 0.6988,
      "step": 3985
    },
    {
      "epoch": 1.6100701984748245,
      "grad_norm": 0.4375,
      "learning_rate": 0.0002037621359223301,
      "loss": 0.5717,
      "step": 3986
    },
    {
      "epoch": 1.6104742184738146,
      "grad_norm": 0.44140625,
      "learning_rate": 0.00020373786407766988,
      "loss": 0.5895,
      "step": 3987
    },
    {
      "epoch": 1.6108782384728044,
      "grad_norm": 0.46875,
      "learning_rate": 0.00020371359223300968,
      "loss": 0.6481,
      "step": 3988
    },
    {
      "epoch": 1.6112822584717943,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0002036893203883495,
      "loss": 0.6022,
      "step": 3989
    },
    {
      "epoch": 1.6116862784707844,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00020366504854368931,
      "loss": 0.6123,
      "step": 3990
    },
    {
      "epoch": 1.6120902984697743,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002036407766990291,
      "loss": 0.6326,
      "step": 3991
    },
    {
      "epoch": 1.6124943184687641,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00020361650485436892,
      "loss": 0.6212,
      "step": 3992
    },
    {
      "epoch": 1.6128983384677542,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00020359223300970872,
      "loss": 0.6145,
      "step": 3993
    },
    {
      "epoch": 1.613302358466744,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0002035679611650485,
      "loss": 0.6237,
      "step": 3994
    },
    {
      "epoch": 1.613706378465734,
      "grad_norm": 0.59375,
      "learning_rate": 0.00020354368932038833,
      "loss": 0.6693,
      "step": 3995
    },
    {
      "epoch": 1.614110398464724,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00020351941747572813,
      "loss": 0.6278,
      "step": 3996
    },
    {
      "epoch": 1.614514418463714,
      "grad_norm": 0.439453125,
      "learning_rate": 0.00020349514563106796,
      "loss": 0.5634,
      "step": 3997
    },
    {
      "epoch": 1.6149184384627038,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00020347087378640776,
      "loss": 0.6013,
      "step": 3998
    },
    {
      "epoch": 1.6153224584616939,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00020344660194174754,
      "loss": 0.6692,
      "step": 3999
    },
    {
      "epoch": 1.615726478460684,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00020342233009708736,
      "loss": 0.6191,
      "step": 4000
    },
    {
      "epoch": 1.6161304984596736,
      "grad_norm": 0.59375,
      "learning_rate": 0.00020339805825242717,
      "loss": 0.7013,
      "step": 4001
    },
    {
      "epoch": 1.6165345184586637,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00020337378640776697,
      "loss": 0.59,
      "step": 4002
    },
    {
      "epoch": 1.6169385384576538,
      "grad_norm": 0.5,
      "learning_rate": 0.0002033495145631068,
      "loss": 0.6179,
      "step": 4003
    },
    {
      "epoch": 1.6173425584566437,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00020332524271844657,
      "loss": 0.6925,
      "step": 4004
    },
    {
      "epoch": 1.6177465784556335,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0002033009708737864,
      "loss": 0.6969,
      "step": 4005
    },
    {
      "epoch": 1.6181505984546236,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002032766990291262,
      "loss": 0.5476,
      "step": 4006
    },
    {
      "epoch": 1.6185546184536135,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00020325242718446598,
      "loss": 0.6579,
      "step": 4007
    },
    {
      "epoch": 1.6189586384526033,
      "grad_norm": 0.59375,
      "learning_rate": 0.0002032281553398058,
      "loss": 0.6946,
      "step": 4008
    },
    {
      "epoch": 1.6193626584515934,
      "grad_norm": 0.5625,
      "learning_rate": 0.0002032038834951456,
      "loss": 0.6262,
      "step": 4009
    },
    {
      "epoch": 1.6197666784505833,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00020317961165048542,
      "loss": 0.6049,
      "step": 4010
    },
    {
      "epoch": 1.6201706984495732,
      "grad_norm": 0.5625,
      "learning_rate": 0.00020315533980582524,
      "loss": 0.6093,
      "step": 4011
    },
    {
      "epoch": 1.6205747184485633,
      "grad_norm": 0.443359375,
      "learning_rate": 0.00020313106796116502,
      "loss": 0.57,
      "step": 4012
    },
    {
      "epoch": 1.6209787384475531,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00020310679611650482,
      "loss": 0.6637,
      "step": 4013
    },
    {
      "epoch": 1.621382758446543,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00020308252427184465,
      "loss": 0.6427,
      "step": 4014
    },
    {
      "epoch": 1.621786778445533,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00020305825242718445,
      "loss": 0.5973,
      "step": 4015
    },
    {
      "epoch": 1.622190798444523,
      "grad_norm": 0.41796875,
      "learning_rate": 0.00020303398058252426,
      "loss": 0.5773,
      "step": 4016
    },
    {
      "epoch": 1.6225948184435128,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00020300970873786406,
      "loss": 0.6271,
      "step": 4017
    },
    {
      "epoch": 1.622998838442503,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020298543689320386,
      "loss": 0.5433,
      "step": 4018
    },
    {
      "epoch": 1.623402858441493,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0002029611650485437,
      "loss": 0.5935,
      "step": 4019
    },
    {
      "epoch": 1.6238068784404827,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00020293689320388347,
      "loss": 0.6409,
      "step": 4020
    },
    {
      "epoch": 1.6242108984394727,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00020291262135922327,
      "loss": 0.6279,
      "step": 4021
    },
    {
      "epoch": 1.6246149184384628,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0002028883495145631,
      "loss": 0.6356,
      "step": 4022
    },
    {
      "epoch": 1.6250189384374527,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0002028640776699029,
      "loss": 0.7348,
      "step": 4023
    },
    {
      "epoch": 1.6254229584364426,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00020283980582524267,
      "loss": 0.5765,
      "step": 4024
    },
    {
      "epoch": 1.6258269784354327,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0002028155339805825,
      "loss": 0.6259,
      "step": 4025
    },
    {
      "epoch": 1.6262309984344225,
      "grad_norm": 0.68359375,
      "learning_rate": 0.0002027912621359223,
      "loss": 0.764,
      "step": 4026
    },
    {
      "epoch": 1.6266350184334124,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00020276699029126214,
      "loss": 0.6181,
      "step": 4027
    },
    {
      "epoch": 1.6270390384324025,
      "grad_norm": 0.65625,
      "learning_rate": 0.00020274271844660194,
      "loss": 0.6367,
      "step": 4028
    },
    {
      "epoch": 1.6274430584313924,
      "grad_norm": 0.625,
      "learning_rate": 0.00020271844660194171,
      "loss": 0.6025,
      "step": 4029
    },
    {
      "epoch": 1.6278470784303822,
      "grad_norm": 0.46875,
      "learning_rate": 0.00020269417475728154,
      "loss": 0.5981,
      "step": 4030
    },
    {
      "epoch": 1.6282510984293723,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020266990291262135,
      "loss": 0.5751,
      "step": 4031
    },
    {
      "epoch": 1.6286551184283622,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00020264563106796112,
      "loss": 0.7564,
      "step": 4032
    },
    {
      "epoch": 1.629059138427352,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00020262135922330095,
      "loss": 0.5732,
      "step": 4033
    },
    {
      "epoch": 1.6294631584263422,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00020259708737864075,
      "loss": 0.6683,
      "step": 4034
    },
    {
      "epoch": 1.629867178425332,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00020257281553398058,
      "loss": 0.5864,
      "step": 4035
    },
    {
      "epoch": 1.6302711984243219,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020254854368932038,
      "loss": 0.5603,
      "step": 4036
    },
    {
      "epoch": 1.630675218423312,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00020252427184466016,
      "loss": 0.636,
      "step": 4037
    },
    {
      "epoch": 1.631079238422302,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002025,
      "loss": 0.5818,
      "step": 4038
    },
    {
      "epoch": 1.6314832584212917,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002024757281553398,
      "loss": 0.6539,
      "step": 4039
    },
    {
      "epoch": 1.6318872784202818,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002024514563106796,
      "loss": 0.5583,
      "step": 4040
    },
    {
      "epoch": 1.632291298419272,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020242718446601942,
      "loss": 0.6434,
      "step": 4041
    },
    {
      "epoch": 1.6326953184182615,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002024029126213592,
      "loss": 0.6367,
      "step": 4042
    },
    {
      "epoch": 1.6330993384172516,
      "grad_norm": 0.50390625,
      "learning_rate": 0.000202378640776699,
      "loss": 0.6103,
      "step": 4043
    },
    {
      "epoch": 1.6335033584162417,
      "grad_norm": 0.44140625,
      "learning_rate": 0.00020235436893203883,
      "loss": 0.6087,
      "step": 4044
    },
    {
      "epoch": 1.6339073784152316,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0002023300970873786,
      "loss": 0.6074,
      "step": 4045
    },
    {
      "epoch": 1.6343113984142215,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00020230582524271843,
      "loss": 0.6868,
      "step": 4046
    },
    {
      "epoch": 1.6347154184132116,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00020228155339805824,
      "loss": 0.6379,
      "step": 4047
    },
    {
      "epoch": 1.6351194384122014,
      "grad_norm": 0.427734375,
      "learning_rate": 0.00020225728155339804,
      "loss": 0.6029,
      "step": 4048
    },
    {
      "epoch": 1.6355234584111913,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00020223300970873787,
      "loss": 0.6393,
      "step": 4049
    },
    {
      "epoch": 1.6359274784101814,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00020220873786407764,
      "loss": 0.5569,
      "step": 4050
    },
    {
      "epoch": 1.6363314984091712,
      "grad_norm": 0.625,
      "learning_rate": 0.00020218446601941745,
      "loss": 0.6752,
      "step": 4051
    },
    {
      "epoch": 1.6367355184081611,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00020216019417475728,
      "loss": 0.5866,
      "step": 4052
    },
    {
      "epoch": 1.6371395384071512,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00020213592233009708,
      "loss": 0.6311,
      "step": 4053
    },
    {
      "epoch": 1.637543558406141,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00020211165048543685,
      "loss": 0.5931,
      "step": 4054
    },
    {
      "epoch": 1.637947578405131,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00020208737864077668,
      "loss": 0.5893,
      "step": 4055
    },
    {
      "epoch": 1.638351598404121,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00020206310679611649,
      "loss": 0.5986,
      "step": 4056
    },
    {
      "epoch": 1.638755618403111,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00020203883495145631,
      "loss": 0.6039,
      "step": 4057
    },
    {
      "epoch": 1.6391596384021008,
      "grad_norm": 0.546875,
      "learning_rate": 0.0002020145631067961,
      "loss": 0.6164,
      "step": 4058
    },
    {
      "epoch": 1.6395636584010909,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0002019902912621359,
      "loss": 0.6531,
      "step": 4059
    },
    {
      "epoch": 1.639967678400081,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020196601941747572,
      "loss": 0.614,
      "step": 4060
    },
    {
      "epoch": 1.6403716983990706,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00020194174757281552,
      "loss": 0.6463,
      "step": 4061
    },
    {
      "epoch": 1.6407757183980607,
      "grad_norm": 0.47265625,
      "learning_rate": 0.0002019174757281553,
      "loss": 0.6887,
      "step": 4062
    },
    {
      "epoch": 1.6411797383970508,
      "grad_norm": 0.416015625,
      "learning_rate": 0.00020189320388349513,
      "loss": 0.5982,
      "step": 4063
    },
    {
      "epoch": 1.6415837583960406,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00020186893203883493,
      "loss": 0.7205,
      "step": 4064
    },
    {
      "epoch": 1.6419877783950305,
      "grad_norm": 0.578125,
      "learning_rate": 0.00020184466019417473,
      "loss": 0.6422,
      "step": 4065
    },
    {
      "epoch": 1.6423917983940206,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00020182038834951456,
      "loss": 0.5709,
      "step": 4066
    },
    {
      "epoch": 1.6427958183930105,
      "grad_norm": 0.546875,
      "learning_rate": 0.00020179611650485434,
      "loss": 0.6591,
      "step": 4067
    },
    {
      "epoch": 1.6431998383920003,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00020177184466019417,
      "loss": 0.6177,
      "step": 4068
    },
    {
      "epoch": 1.6436038583909904,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00020174757281553397,
      "loss": 0.6671,
      "step": 4069
    },
    {
      "epoch": 1.6440078783899803,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020172330097087375,
      "loss": 0.6676,
      "step": 4070
    },
    {
      "epoch": 1.6444118983889702,
      "grad_norm": 0.59375,
      "learning_rate": 0.00020169902912621357,
      "loss": 0.6786,
      "step": 4071
    },
    {
      "epoch": 1.6448159183879603,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00020167475728155338,
      "loss": 0.7389,
      "step": 4072
    },
    {
      "epoch": 1.6452199383869501,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00020165048543689318,
      "loss": 0.6544,
      "step": 4073
    },
    {
      "epoch": 1.64562395838594,
      "grad_norm": 0.5078125,
      "learning_rate": 0.000201626213592233,
      "loss": 0.5852,
      "step": 4074
    },
    {
      "epoch": 1.64602797838493,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00020160194174757278,
      "loss": 0.6623,
      "step": 4075
    },
    {
      "epoch": 1.64643199838392,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0002015776699029126,
      "loss": 0.7391,
      "step": 4076
    },
    {
      "epoch": 1.6468360183829098,
      "grad_norm": 0.546875,
      "learning_rate": 0.00020155339805825242,
      "loss": 0.6432,
      "step": 4077
    },
    {
      "epoch": 1.6472400383819,
      "grad_norm": 0.546875,
      "learning_rate": 0.00020152912621359222,
      "loss": 0.6078,
      "step": 4078
    },
    {
      "epoch": 1.64764405838089,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00020150485436893202,
      "loss": 0.6073,
      "step": 4079
    },
    {
      "epoch": 1.6480480783798797,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00020148058252427182,
      "loss": 0.6553,
      "step": 4080
    },
    {
      "epoch": 1.6484520983788697,
      "grad_norm": 0.46875,
      "learning_rate": 0.00020145631067961163,
      "loss": 0.609,
      "step": 4081
    },
    {
      "epoch": 1.6488561183778598,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00020143203883495145,
      "loss": 0.6865,
      "step": 4082
    },
    {
      "epoch": 1.6492601383768497,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00020140776699029123,
      "loss": 0.6918,
      "step": 4083
    },
    {
      "epoch": 1.6496641583758396,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020138349514563103,
      "loss": 0.6827,
      "step": 4084
    },
    {
      "epoch": 1.6500681783748297,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020135922330097086,
      "loss": 0.6366,
      "step": 4085
    },
    {
      "epoch": 1.6504721983738195,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00020133495145631066,
      "loss": 0.676,
      "step": 4086
    },
    {
      "epoch": 1.6508762183728094,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002013106796116505,
      "loss": 0.6307,
      "step": 4087
    },
    {
      "epoch": 1.6512802383717995,
      "grad_norm": 0.5,
      "learning_rate": 0.00020128640776699027,
      "loss": 0.6253,
      "step": 4088
    },
    {
      "epoch": 1.6516842583707894,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020126213592233007,
      "loss": 0.6758,
      "step": 4089
    },
    {
      "epoch": 1.6520882783697792,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002012378640776699,
      "loss": 0.6228,
      "step": 4090
    },
    {
      "epoch": 1.6524922983687693,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0002012135922330097,
      "loss": 0.7523,
      "step": 4091
    },
    {
      "epoch": 1.6528963183677592,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00020118932038834948,
      "loss": 0.591,
      "step": 4092
    },
    {
      "epoch": 1.653300338366749,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0002011650485436893,
      "loss": 0.6281,
      "step": 4093
    },
    {
      "epoch": 1.6537043583657391,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0002011407766990291,
      "loss": 0.656,
      "step": 4094
    },
    {
      "epoch": 1.654108378364729,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00020111650485436889,
      "loss": 0.683,
      "step": 4095
    },
    {
      "epoch": 1.6545123983637189,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00020109223300970871,
      "loss": 0.6874,
      "step": 4096
    },
    {
      "epoch": 1.654916418362709,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00020106796116504852,
      "loss": 0.6464,
      "step": 4097
    },
    {
      "epoch": 1.655320438361699,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00020104368932038835,
      "loss": 0.6508,
      "step": 4098
    },
    {
      "epoch": 1.6557244583606887,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00020101941747572815,
      "loss": 0.5731,
      "step": 4099
    },
    {
      "epoch": 1.6561284783596788,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00020099514563106792,
      "loss": 0.5883,
      "step": 4100
    },
    {
      "epoch": 1.656532498358669,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00020097087378640775,
      "loss": 0.6026,
      "step": 4101
    },
    {
      "epoch": 1.6569365183576585,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00020094660194174756,
      "loss": 0.6133,
      "step": 4102
    },
    {
      "epoch": 1.6573405383566486,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020092233009708736,
      "loss": 0.6144,
      "step": 4103
    },
    {
      "epoch": 1.6577445583556387,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002008980582524272,
      "loss": 0.6164,
      "step": 4104
    },
    {
      "epoch": 1.6581485783546286,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00020087378640776696,
      "loss": 0.5736,
      "step": 4105
    },
    {
      "epoch": 1.6585525983536185,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0002008495145631068,
      "loss": 0.6586,
      "step": 4106
    },
    {
      "epoch": 1.6589566183526085,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0002008252427184466,
      "loss": 0.5987,
      "step": 4107
    },
    {
      "epoch": 1.6593606383515984,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00020080097087378637,
      "loss": 0.574,
      "step": 4108
    },
    {
      "epoch": 1.6597646583505883,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0002007766990291262,
      "loss": 0.6166,
      "step": 4109
    },
    {
      "epoch": 1.6601686783495784,
      "grad_norm": 0.53125,
      "learning_rate": 0.000200752427184466,
      "loss": 0.5908,
      "step": 4110
    },
    {
      "epoch": 1.6605726983485682,
      "grad_norm": 0.48046875,
      "learning_rate": 0.0002007281553398058,
      "loss": 0.5935,
      "step": 4111
    },
    {
      "epoch": 1.6609767183475581,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00020070388349514563,
      "loss": 0.6019,
      "step": 4112
    },
    {
      "epoch": 1.6613807383465482,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0002006796116504854,
      "loss": 0.5827,
      "step": 4113
    },
    {
      "epoch": 1.661784758345538,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0002006553398058252,
      "loss": 0.6532,
      "step": 4114
    },
    {
      "epoch": 1.662188778344528,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00020063106796116504,
      "loss": 0.5974,
      "step": 4115
    },
    {
      "epoch": 1.662592798343518,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00020060679611650484,
      "loss": 0.7018,
      "step": 4116
    },
    {
      "epoch": 1.662996818342508,
      "grad_norm": 0.5625,
      "learning_rate": 0.00020058252427184464,
      "loss": 0.5909,
      "step": 4117
    },
    {
      "epoch": 1.6634008383414978,
      "grad_norm": 0.609375,
      "learning_rate": 0.00020055825242718445,
      "loss": 0.6562,
      "step": 4118
    },
    {
      "epoch": 1.6638048583404879,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020053398058252425,
      "loss": 0.606,
      "step": 4119
    },
    {
      "epoch": 1.664208878339478,
      "grad_norm": 0.609375,
      "learning_rate": 0.00020050970873786408,
      "loss": 0.6799,
      "step": 4120
    },
    {
      "epoch": 1.6646128983384676,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00020048543689320385,
      "loss": 0.6155,
      "step": 4121
    },
    {
      "epoch": 1.6650169183374577,
      "grad_norm": 0.439453125,
      "learning_rate": 0.00020046116504854366,
      "loss": 0.6144,
      "step": 4122
    },
    {
      "epoch": 1.6654209383364478,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00020043689320388349,
      "loss": 0.6098,
      "step": 4123
    },
    {
      "epoch": 1.6658249583354376,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0002004126213592233,
      "loss": 0.5577,
      "step": 4124
    },
    {
      "epoch": 1.6662289783344275,
      "grad_norm": 0.59375,
      "learning_rate": 0.00020038834951456306,
      "loss": 0.679,
      "step": 4125
    },
    {
      "epoch": 1.6666329983334176,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0002003640776699029,
      "loss": 0.6668,
      "step": 4126
    },
    {
      "epoch": 1.6670370183324075,
      "grad_norm": 0.53125,
      "learning_rate": 0.0002003398058252427,
      "loss": 0.7131,
      "step": 4127
    },
    {
      "epoch": 1.6674410383313973,
      "grad_norm": 0.5625,
      "learning_rate": 0.00020031553398058252,
      "loss": 0.6416,
      "step": 4128
    },
    {
      "epoch": 1.6678450583303874,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00020029126213592233,
      "loss": 0.6659,
      "step": 4129
    },
    {
      "epoch": 1.6682490783293773,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0002002669902912621,
      "loss": 0.6144,
      "step": 4130
    },
    {
      "epoch": 1.6686530983283672,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00020024271844660193,
      "loss": 0.5277,
      "step": 4131
    },
    {
      "epoch": 1.6690571183273573,
      "grad_norm": 0.5625,
      "learning_rate": 0.00020021844660194173,
      "loss": 0.6558,
      "step": 4132
    },
    {
      "epoch": 1.6694611383263471,
      "grad_norm": 0.443359375,
      "learning_rate": 0.0002001941747572815,
      "loss": 0.6127,
      "step": 4133
    },
    {
      "epoch": 1.669865158325337,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020016990291262134,
      "loss": 0.6383,
      "step": 4134
    },
    {
      "epoch": 1.670269178324327,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00020014563106796114,
      "loss": 0.6041,
      "step": 4135
    },
    {
      "epoch": 1.670673198323317,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00020012135922330097,
      "loss": 0.6077,
      "step": 4136
    },
    {
      "epoch": 1.6710772183223068,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00020009708737864077,
      "loss": 0.6192,
      "step": 4137
    },
    {
      "epoch": 1.671481238321297,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00020007281553398055,
      "loss": 0.6973,
      "step": 4138
    },
    {
      "epoch": 1.671885258320287,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00020004854368932038,
      "loss": 0.584,
      "step": 4139
    },
    {
      "epoch": 1.6722892783192767,
      "grad_norm": 0.515625,
      "learning_rate": 0.00020002427184466018,
      "loss": 0.6627,
      "step": 4140
    },
    {
      "epoch": 1.6726932983182667,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00019999999999999998,
      "loss": 0.5462,
      "step": 4141
    },
    {
      "epoch": 1.6730973183172568,
      "grad_norm": 0.578125,
      "learning_rate": 0.00019997572815533978,
      "loss": 0.7097,
      "step": 4142
    },
    {
      "epoch": 1.6735013383162467,
      "grad_norm": 0.46484375,
      "learning_rate": 0.0001999514563106796,
      "loss": 0.5178,
      "step": 4143
    },
    {
      "epoch": 1.6739053583152366,
      "grad_norm": 0.451171875,
      "learning_rate": 0.0001999271844660194,
      "loss": 0.6051,
      "step": 4144
    },
    {
      "epoch": 1.6743093783142267,
      "grad_norm": 0.43359375,
      "learning_rate": 0.00019990291262135922,
      "loss": 0.5271,
      "step": 4145
    },
    {
      "epoch": 1.6747133983132165,
      "grad_norm": 0.65234375,
      "learning_rate": 0.000199878640776699,
      "loss": 0.6586,
      "step": 4146
    },
    {
      "epoch": 1.6751174183122064,
      "grad_norm": 0.453125,
      "learning_rate": 0.00019985436893203882,
      "loss": 0.6023,
      "step": 4147
    },
    {
      "epoch": 1.6755214383111965,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019983009708737863,
      "loss": 0.6284,
      "step": 4148
    },
    {
      "epoch": 1.6759254583101864,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00019980582524271843,
      "loss": 0.5607,
      "step": 4149
    },
    {
      "epoch": 1.6763294783091762,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00019978155339805826,
      "loss": 0.6816,
      "step": 4150
    },
    {
      "epoch": 1.6767334983081663,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00019975728155339803,
      "loss": 0.5999,
      "step": 4151
    },
    {
      "epoch": 1.6771375183071562,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00019973300970873784,
      "loss": 0.5757,
      "step": 4152
    },
    {
      "epoch": 1.677541538306146,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00019970873786407766,
      "loss": 0.6697,
      "step": 4153
    },
    {
      "epoch": 1.6779455583051361,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00019968446601941747,
      "loss": 0.605,
      "step": 4154
    },
    {
      "epoch": 1.678349578304126,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00019966019417475724,
      "loss": 0.5784,
      "step": 4155
    },
    {
      "epoch": 1.6787535983031159,
      "grad_norm": 0.546875,
      "learning_rate": 0.00019963592233009707,
      "loss": 0.6555,
      "step": 4156
    },
    {
      "epoch": 1.679157618302106,
      "grad_norm": 0.431640625,
      "learning_rate": 0.00019961165048543687,
      "loss": 0.5971,
      "step": 4157
    },
    {
      "epoch": 1.679561638301096,
      "grad_norm": 0.47265625,
      "learning_rate": 0.0001995873786407767,
      "loss": 0.6121,
      "step": 4158
    },
    {
      "epoch": 1.6799656583000857,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00019956310679611648,
      "loss": 0.6157,
      "step": 4159
    },
    {
      "epoch": 1.6803696782990758,
      "grad_norm": 0.609375,
      "learning_rate": 0.00019953883495145628,
      "loss": 0.6872,
      "step": 4160
    },
    {
      "epoch": 1.6807736982980659,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0001995145631067961,
      "loss": 0.511,
      "step": 4161
    },
    {
      "epoch": 1.6811777182970558,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0001994902912621359,
      "loss": 0.5778,
      "step": 4162
    },
    {
      "epoch": 1.6815817382960456,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0001994660194174757,
      "loss": 0.6096,
      "step": 4163
    },
    {
      "epoch": 1.6819857582950357,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00019944174757281552,
      "loss": 0.662,
      "step": 4164
    },
    {
      "epoch": 1.6823897782940256,
      "grad_norm": 0.625,
      "learning_rate": 0.00019941747572815532,
      "loss": 0.6532,
      "step": 4165
    },
    {
      "epoch": 1.6827937982930155,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00019939320388349512,
      "loss": 0.7075,
      "step": 4166
    },
    {
      "epoch": 1.6831978182920055,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00019936893203883495,
      "loss": 0.6672,
      "step": 4167
    },
    {
      "epoch": 1.6836018382909954,
      "grad_norm": 0.75390625,
      "learning_rate": 0.00019934466019417473,
      "loss": 0.7211,
      "step": 4168
    },
    {
      "epoch": 1.6840058582899853,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00019932038834951456,
      "loss": 0.6628,
      "step": 4169
    },
    {
      "epoch": 1.6844098782889754,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019929611650485436,
      "loss": 0.5906,
      "step": 4170
    },
    {
      "epoch": 1.6848138982879652,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019927184466019413,
      "loss": 0.656,
      "step": 4171
    },
    {
      "epoch": 1.685217918286955,
      "grad_norm": 0.453125,
      "learning_rate": 0.00019924757281553396,
      "loss": 0.5852,
      "step": 4172
    },
    {
      "epoch": 1.6856219382859452,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00019922330097087377,
      "loss": 0.6146,
      "step": 4173
    },
    {
      "epoch": 1.686025958284935,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019919902912621357,
      "loss": 0.6353,
      "step": 4174
    },
    {
      "epoch": 1.686429978283925,
      "grad_norm": 0.578125,
      "learning_rate": 0.0001991747572815534,
      "loss": 0.635,
      "step": 4175
    },
    {
      "epoch": 1.686833998282915,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00019915048543689317,
      "loss": 0.6439,
      "step": 4176
    },
    {
      "epoch": 1.6872380182819051,
      "grad_norm": 0.59375,
      "learning_rate": 0.000199126213592233,
      "loss": 0.6345,
      "step": 4177
    },
    {
      "epoch": 1.6876420382808948,
      "grad_norm": 0.546875,
      "learning_rate": 0.0001991019417475728,
      "loss": 0.717,
      "step": 4178
    },
    {
      "epoch": 1.6880460582798849,
      "grad_norm": 0.609375,
      "learning_rate": 0.0001990776699029126,
      "loss": 0.652,
      "step": 4179
    },
    {
      "epoch": 1.688450078278875,
      "grad_norm": 0.5625,
      "learning_rate": 0.0001990533980582524,
      "loss": 0.7065,
      "step": 4180
    },
    {
      "epoch": 1.6888540982778646,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0001990291262135922,
      "loss": 0.6722,
      "step": 4181
    },
    {
      "epoch": 1.6892581182768547,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00019900485436893201,
      "loss": 0.5743,
      "step": 4182
    },
    {
      "epoch": 1.6896621382758448,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00019898058252427184,
      "loss": 0.7896,
      "step": 4183
    },
    {
      "epoch": 1.6900661582748346,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00019895631067961162,
      "loss": 0.6326,
      "step": 4184
    },
    {
      "epoch": 1.6904701782738245,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019893203883495142,
      "loss": 0.6492,
      "step": 4185
    },
    {
      "epoch": 1.6908741982728146,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00019890776699029125,
      "loss": 0.6449,
      "step": 4186
    },
    {
      "epoch": 1.6912782182718045,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00019888349514563105,
      "loss": 0.6816,
      "step": 4187
    },
    {
      "epoch": 1.6916822382707943,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00019885922330097088,
      "loss": 0.5527,
      "step": 4188
    },
    {
      "epoch": 1.6920862582697844,
      "grad_norm": 6.46875,
      "learning_rate": 0.00019883495145631066,
      "loss": 0.6755,
      "step": 4189
    },
    {
      "epoch": 1.6924902782687743,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019881067961165046,
      "loss": 0.6113,
      "step": 4190
    },
    {
      "epoch": 1.6928942982677642,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0001987864077669903,
      "loss": 0.5618,
      "step": 4191
    },
    {
      "epoch": 1.6932983182667543,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0001987621359223301,
      "loss": 0.6712,
      "step": 4192
    },
    {
      "epoch": 1.6937023382657441,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00019873786407766987,
      "loss": 0.6631,
      "step": 4193
    },
    {
      "epoch": 1.694106358264734,
      "grad_norm": 0.515625,
      "learning_rate": 0.0001987135922330097,
      "loss": 0.6104,
      "step": 4194
    },
    {
      "epoch": 1.694510378263724,
      "grad_norm": 0.41015625,
      "learning_rate": 0.0001986893203883495,
      "loss": 0.5819,
      "step": 4195
    },
    {
      "epoch": 1.694914398262714,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00019866504854368927,
      "loss": 0.654,
      "step": 4196
    },
    {
      "epoch": 1.6953184182617038,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0001986407766990291,
      "loss": 0.5736,
      "step": 4197
    },
    {
      "epoch": 1.695722438260694,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0001986165048543689,
      "loss": 0.6528,
      "step": 4198
    },
    {
      "epoch": 1.696126458259684,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00019859223300970874,
      "loss": 0.4809,
      "step": 4199
    },
    {
      "epoch": 1.6965304782586736,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00019856796116504854,
      "loss": 0.7014,
      "step": 4200
    },
    {
      "epoch": 1.6969344982576637,
      "grad_norm": 0.470703125,
      "learning_rate": 0.0001985436893203883,
      "loss": 0.6123,
      "step": 4201
    },
    {
      "epoch": 1.6973385182566538,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00019851941747572814,
      "loss": 0.6353,
      "step": 4202
    },
    {
      "epoch": 1.6977425382556437,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019849514563106794,
      "loss": 0.604,
      "step": 4203
    },
    {
      "epoch": 1.6981465582546336,
      "grad_norm": 0.447265625,
      "learning_rate": 0.00019847087378640775,
      "loss": 0.5791,
      "step": 4204
    },
    {
      "epoch": 1.6985505782536237,
      "grad_norm": 1.53125,
      "learning_rate": 0.00019844660194174758,
      "loss": 0.6939,
      "step": 4205
    },
    {
      "epoch": 1.6989545982526135,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00019842233009708735,
      "loss": 0.6363,
      "step": 4206
    },
    {
      "epoch": 1.6993586182516034,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00019839805825242718,
      "loss": 0.609,
      "step": 4207
    },
    {
      "epoch": 1.6997626382505935,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00019837378640776698,
      "loss": 0.5936,
      "step": 4208
    },
    {
      "epoch": 1.7001666582495834,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00019834951456310676,
      "loss": 0.6,
      "step": 4209
    },
    {
      "epoch": 1.7005706782485732,
      "grad_norm": 0.470703125,
      "learning_rate": 0.0001983252427184466,
      "loss": 0.6541,
      "step": 4210
    },
    {
      "epoch": 1.7009746982475633,
      "grad_norm": 0.451171875,
      "learning_rate": 0.0001983009708737864,
      "loss": 0.5541,
      "step": 4211
    },
    {
      "epoch": 1.7013787182465532,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0001982766990291262,
      "loss": 0.7211,
      "step": 4212
    },
    {
      "epoch": 1.701782738245543,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00019825242718446602,
      "loss": 0.646,
      "step": 4213
    },
    {
      "epoch": 1.7021867582445331,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0001982281553398058,
      "loss": 0.655,
      "step": 4214
    },
    {
      "epoch": 1.702590778243523,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0001982038834951456,
      "loss": 0.5879,
      "step": 4215
    },
    {
      "epoch": 1.7029947982425129,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00019817961165048543,
      "loss": 0.7266,
      "step": 4216
    },
    {
      "epoch": 1.703398818241503,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019815533980582523,
      "loss": 0.5886,
      "step": 4217
    },
    {
      "epoch": 1.703802838240493,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00019813106796116503,
      "loss": 0.7317,
      "step": 4218
    },
    {
      "epoch": 1.7042068582394827,
      "grad_norm": 0.515625,
      "learning_rate": 0.00019810679611650484,
      "loss": 0.6115,
      "step": 4219
    },
    {
      "epoch": 1.7046108782384728,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00019808252427184464,
      "loss": 0.6268,
      "step": 4220
    },
    {
      "epoch": 1.7050148982374629,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00019805825242718447,
      "loss": 0.5872,
      "step": 4221
    },
    {
      "epoch": 1.7054189182364528,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00019803398058252424,
      "loss": 0.7107,
      "step": 4222
    },
    {
      "epoch": 1.7058229382354426,
      "grad_norm": 0.451171875,
      "learning_rate": 0.00019800970873786405,
      "loss": 0.6167,
      "step": 4223
    },
    {
      "epoch": 1.7062269582344327,
      "grad_norm": 0.515625,
      "learning_rate": 0.00019798543689320387,
      "loss": 0.6443,
      "step": 4224
    },
    {
      "epoch": 1.7066309782334226,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019796116504854368,
      "loss": 0.5877,
      "step": 4225
    },
    {
      "epoch": 1.7070349982324124,
      "grad_norm": 0.53125,
      "learning_rate": 0.00019793689320388345,
      "loss": 0.5694,
      "step": 4226
    },
    {
      "epoch": 1.7074390182314025,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00019791262135922328,
      "loss": 0.5858,
      "step": 4227
    },
    {
      "epoch": 1.7078430382303924,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00019788834951456308,
      "loss": 0.6234,
      "step": 4228
    },
    {
      "epoch": 1.7082470582293823,
      "grad_norm": 0.515625,
      "learning_rate": 0.00019786407766990291,
      "loss": 0.6255,
      "step": 4229
    },
    {
      "epoch": 1.7086510782283724,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00019783980582524272,
      "loss": 0.6396,
      "step": 4230
    },
    {
      "epoch": 1.7090550982273622,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0001978155339805825,
      "loss": 0.6725,
      "step": 4231
    },
    {
      "epoch": 1.709459118226352,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00019779126213592232,
      "loss": 0.5899,
      "step": 4232
    },
    {
      "epoch": 1.7098631382253422,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019776699029126212,
      "loss": 0.6283,
      "step": 4233
    },
    {
      "epoch": 1.710267158224332,
      "grad_norm": 0.53125,
      "learning_rate": 0.0001977427184466019,
      "loss": 0.6149,
      "step": 4234
    },
    {
      "epoch": 1.710671178223322,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019771844660194173,
      "loss": 0.5615,
      "step": 4235
    },
    {
      "epoch": 1.711075198222312,
      "grad_norm": 0.671875,
      "learning_rate": 0.00019769417475728153,
      "loss": 0.7406,
      "step": 4236
    },
    {
      "epoch": 1.7114792182213021,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00019766990291262136,
      "loss": 0.6056,
      "step": 4237
    },
    {
      "epoch": 1.7118832382202918,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019764563106796116,
      "loss": 0.6069,
      "step": 4238
    },
    {
      "epoch": 1.7122872582192818,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00019762135922330094,
      "loss": 0.6153,
      "step": 4239
    },
    {
      "epoch": 1.712691278218272,
      "grad_norm": 0.53125,
      "learning_rate": 0.00019759708737864077,
      "loss": 0.6172,
      "step": 4240
    },
    {
      "epoch": 1.7130952982172618,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00019757281553398057,
      "loss": 0.602,
      "step": 4241
    },
    {
      "epoch": 1.7134993182162517,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00019754854368932037,
      "loss": 0.7025,
      "step": 4242
    },
    {
      "epoch": 1.7139033382152418,
      "grad_norm": 0.515625,
      "learning_rate": 0.00019752427184466017,
      "loss": 0.622,
      "step": 4243
    },
    {
      "epoch": 1.7143073582142316,
      "grad_norm": 0.46875,
      "learning_rate": 0.00019749999999999998,
      "loss": 0.5565,
      "step": 4244
    },
    {
      "epoch": 1.7147113782132215,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00019747572815533978,
      "loss": 0.6998,
      "step": 4245
    },
    {
      "epoch": 1.7151153982122116,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0001974514563106796,
      "loss": 0.6242,
      "step": 4246
    },
    {
      "epoch": 1.7155194182112015,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00019742718446601938,
      "loss": 0.6021,
      "step": 4247
    },
    {
      "epoch": 1.7159234382101913,
      "grad_norm": 0.484375,
      "learning_rate": 0.0001974029126213592,
      "loss": 0.5925,
      "step": 4248
    },
    {
      "epoch": 1.7163274582091814,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00019737864077669901,
      "loss": 0.6809,
      "step": 4249
    },
    {
      "epoch": 1.7167314782081713,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00019735436893203882,
      "loss": 0.6225,
      "step": 4250
    },
    {
      "epoch": 1.7171354982071612,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00019733009708737865,
      "loss": 0.5768,
      "step": 4251
    },
    {
      "epoch": 1.7175395182061513,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00019730582524271842,
      "loss": 0.5596,
      "step": 4252
    },
    {
      "epoch": 1.7179435382051411,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00019728155339805822,
      "loss": 0.5921,
      "step": 4253
    },
    {
      "epoch": 1.718347558204131,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019725728155339805,
      "loss": 0.6899,
      "step": 4254
    },
    {
      "epoch": 1.718751578203121,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00019723300970873786,
      "loss": 0.666,
      "step": 4255
    },
    {
      "epoch": 1.7191555982021112,
      "grad_norm": 0.546875,
      "learning_rate": 0.00019720873786407763,
      "loss": 0.6341,
      "step": 4256
    },
    {
      "epoch": 1.7195596182011008,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00019718446601941746,
      "loss": 0.6192,
      "step": 4257
    },
    {
      "epoch": 1.719963638200091,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00019716019417475726,
      "loss": 0.6355,
      "step": 4258
    },
    {
      "epoch": 1.720367658199081,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0001971359223300971,
      "loss": 0.6137,
      "step": 4259
    },
    {
      "epoch": 1.7207716781980706,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019711165048543687,
      "loss": 0.6713,
      "step": 4260
    },
    {
      "epoch": 1.7211756981970607,
      "grad_norm": 0.443359375,
      "learning_rate": 0.00019708737864077667,
      "loss": 0.5303,
      "step": 4261
    },
    {
      "epoch": 1.7215797181960508,
      "grad_norm": 0.47265625,
      "learning_rate": 0.0001970631067961165,
      "loss": 0.5571,
      "step": 4262
    },
    {
      "epoch": 1.7219837381950407,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0001970388349514563,
      "loss": 0.589,
      "step": 4263
    },
    {
      "epoch": 1.7223877581940306,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00019701456310679608,
      "loss": 0.6458,
      "step": 4264
    },
    {
      "epoch": 1.7227917781930207,
      "grad_norm": 0.458984375,
      "learning_rate": 0.0001969902912621359,
      "loss": 0.598,
      "step": 4265
    },
    {
      "epoch": 1.7231957981920105,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0001969660194174757,
      "loss": 0.6846,
      "step": 4266
    },
    {
      "epoch": 1.7235998181910004,
      "grad_norm": 0.447265625,
      "learning_rate": 0.0001969417475728155,
      "loss": 0.5785,
      "step": 4267
    },
    {
      "epoch": 1.7240038381899905,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00019691747572815534,
      "loss": 0.6694,
      "step": 4268
    },
    {
      "epoch": 1.7244078581889803,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00019689320388349512,
      "loss": 0.5859,
      "step": 4269
    },
    {
      "epoch": 1.7248118781879702,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00019686893203883495,
      "loss": 0.5672,
      "step": 4270
    },
    {
      "epoch": 1.7252158981869603,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00019684466019417475,
      "loss": 0.6107,
      "step": 4271
    },
    {
      "epoch": 1.7256199181859502,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00019682038834951452,
      "loss": 0.6687,
      "step": 4272
    },
    {
      "epoch": 1.72602393818494,
      "grad_norm": 0.546875,
      "learning_rate": 0.00019679611650485435,
      "loss": 0.5938,
      "step": 4273
    },
    {
      "epoch": 1.7264279581839301,
      "grad_norm": 0.431640625,
      "learning_rate": 0.00019677184466019415,
      "loss": 0.5936,
      "step": 4274
    },
    {
      "epoch": 1.72683197818292,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00019674757281553396,
      "loss": 0.6218,
      "step": 4275
    },
    {
      "epoch": 1.7272359981819099,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019672330097087379,
      "loss": 0.6352,
      "step": 4276
    },
    {
      "epoch": 1.7276400181809,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019669902912621356,
      "loss": 0.6044,
      "step": 4277
    },
    {
      "epoch": 1.72804403817989,
      "grad_norm": 0.4609375,
      "learning_rate": 0.0001966747572815534,
      "loss": 0.6511,
      "step": 4278
    },
    {
      "epoch": 1.7284480581788797,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0001966504854368932,
      "loss": 0.6773,
      "step": 4279
    },
    {
      "epoch": 1.7288520781778698,
      "grad_norm": 0.5,
      "learning_rate": 0.000196626213592233,
      "loss": 0.5552,
      "step": 4280
    },
    {
      "epoch": 1.7292560981768599,
      "grad_norm": 0.47265625,
      "learning_rate": 0.0001966019417475728,
      "loss": 0.5321,
      "step": 4281
    },
    {
      "epoch": 1.7296601181758497,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0001965776699029126,
      "loss": 0.6235,
      "step": 4282
    },
    {
      "epoch": 1.7300641381748396,
      "grad_norm": 0.45703125,
      "learning_rate": 0.0001965533980582524,
      "loss": 0.5795,
      "step": 4283
    },
    {
      "epoch": 1.7304681581738297,
      "grad_norm": 0.5625,
      "learning_rate": 0.00019652912621359223,
      "loss": 0.6645,
      "step": 4284
    },
    {
      "epoch": 1.7308721781728196,
      "grad_norm": 0.57421875,
      "learning_rate": 0.000196504854368932,
      "loss": 0.5845,
      "step": 4285
    },
    {
      "epoch": 1.7312761981718094,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0001964805825242718,
      "loss": 0.6175,
      "step": 4286
    },
    {
      "epoch": 1.7316802181707995,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019645631067961164,
      "loss": 0.6152,
      "step": 4287
    },
    {
      "epoch": 1.7320842381697894,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019643203883495144,
      "loss": 0.6025,
      "step": 4288
    },
    {
      "epoch": 1.7324882581687793,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00019640776699029127,
      "loss": 0.5097,
      "step": 4289
    },
    {
      "epoch": 1.7328922781677694,
      "grad_norm": 0.546875,
      "learning_rate": 0.00019638349514563105,
      "loss": 0.5831,
      "step": 4290
    },
    {
      "epoch": 1.7332962981667592,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00019635922330097085,
      "loss": 0.6224,
      "step": 4291
    },
    {
      "epoch": 1.733700318165749,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00019633495145631068,
      "loss": 0.6417,
      "step": 4292
    },
    {
      "epoch": 1.7341043381647392,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00019631067961165048,
      "loss": 0.6879,
      "step": 4293
    },
    {
      "epoch": 1.734508358163729,
      "grad_norm": 0.59375,
      "learning_rate": 0.00019628640776699026,
      "loss": 0.6752,
      "step": 4294
    },
    {
      "epoch": 1.734912378162719,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00019626213592233009,
      "loss": 0.5898,
      "step": 4295
    },
    {
      "epoch": 1.735316398161709,
      "grad_norm": 0.625,
      "learning_rate": 0.0001962378640776699,
      "loss": 0.6029,
      "step": 4296
    },
    {
      "epoch": 1.735720418160699,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00019621359223300966,
      "loss": 0.5708,
      "step": 4297
    },
    {
      "epoch": 1.7361244381596888,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0001961893203883495,
      "loss": 0.6672,
      "step": 4298
    },
    {
      "epoch": 1.7365284581586788,
      "grad_norm": 0.59375,
      "learning_rate": 0.0001961650485436893,
      "loss": 0.7009,
      "step": 4299
    },
    {
      "epoch": 1.736932478157669,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019614077669902912,
      "loss": 0.6662,
      "step": 4300
    },
    {
      "epoch": 1.7373364981566588,
      "grad_norm": 0.515625,
      "learning_rate": 0.00019611650485436893,
      "loss": 0.5905,
      "step": 4301
    },
    {
      "epoch": 1.7377405181556487,
      "grad_norm": 0.4296875,
      "learning_rate": 0.0001960922330097087,
      "loss": 0.5131,
      "step": 4302
    },
    {
      "epoch": 1.7381445381546388,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00019606796116504853,
      "loss": 0.5799,
      "step": 4303
    },
    {
      "epoch": 1.7385485581536286,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00019604368932038833,
      "loss": 0.5772,
      "step": 4304
    },
    {
      "epoch": 1.7389525781526185,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019601941747572814,
      "loss": 0.6366,
      "step": 4305
    },
    {
      "epoch": 1.7393565981516086,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00019599514563106794,
      "loss": 0.6162,
      "step": 4306
    },
    {
      "epoch": 1.7397606181505985,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00019597087378640774,
      "loss": 0.5966,
      "step": 4307
    },
    {
      "epoch": 1.7401646381495883,
      "grad_norm": 0.5,
      "learning_rate": 0.00019594660194174757,
      "loss": 0.592,
      "step": 4308
    },
    {
      "epoch": 1.7405686581485784,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00019592233009708737,
      "loss": 0.5984,
      "step": 4309
    },
    {
      "epoch": 1.7409726781475683,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00019589805825242715,
      "loss": 0.6118,
      "step": 4310
    },
    {
      "epoch": 1.7413766981465582,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00019587378640776698,
      "loss": 0.6164,
      "step": 4311
    },
    {
      "epoch": 1.7417807181455482,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019584951456310678,
      "loss": 0.7055,
      "step": 4312
    },
    {
      "epoch": 1.7421847381445381,
      "grad_norm": 0.5625,
      "learning_rate": 0.00019582524271844658,
      "loss": 0.5721,
      "step": 4313
    },
    {
      "epoch": 1.742588758143528,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0001958009708737864,
      "loss": 0.7131,
      "step": 4314
    },
    {
      "epoch": 1.742992778142518,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00019577669902912619,
      "loss": 0.6004,
      "step": 4315
    },
    {
      "epoch": 1.7433967981415082,
      "grad_norm": 0.59375,
      "learning_rate": 0.000195752427184466,
      "loss": 0.5967,
      "step": 4316
    },
    {
      "epoch": 1.7438008181404978,
      "grad_norm": 0.447265625,
      "learning_rate": 0.00019572815533980582,
      "loss": 0.5855,
      "step": 4317
    },
    {
      "epoch": 1.744204838139488,
      "grad_norm": 0.5625,
      "learning_rate": 0.00019570388349514562,
      "loss": 0.6388,
      "step": 4318
    },
    {
      "epoch": 1.744608858138478,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00019567961165048542,
      "loss": 0.686,
      "step": 4319
    },
    {
      "epoch": 1.7450128781374679,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00019565533980582522,
      "loss": 0.6187,
      "step": 4320
    },
    {
      "epoch": 1.7454168981364577,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019563106796116503,
      "loss": 0.6163,
      "step": 4321
    },
    {
      "epoch": 1.7458209181354478,
      "grad_norm": 0.625,
      "learning_rate": 0.00019560679611650486,
      "loss": 0.6442,
      "step": 4322
    },
    {
      "epoch": 1.7462249381344377,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00019558252427184463,
      "loss": 0.6387,
      "step": 4323
    },
    {
      "epoch": 1.7466289581334276,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00019555825242718443,
      "loss": 0.6395,
      "step": 4324
    },
    {
      "epoch": 1.7470329781324176,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019553398058252426,
      "loss": 0.6259,
      "step": 4325
    },
    {
      "epoch": 1.7474369981314075,
      "grad_norm": 0.5,
      "learning_rate": 0.00019550970873786407,
      "loss": 0.5683,
      "step": 4326
    },
    {
      "epoch": 1.7478410181303974,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00019548543689320384,
      "loss": 0.5691,
      "step": 4327
    },
    {
      "epoch": 1.7482450381293875,
      "grad_norm": 0.46875,
      "learning_rate": 0.00019546116504854367,
      "loss": 0.5573,
      "step": 4328
    },
    {
      "epoch": 1.7486490581283773,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00019543689320388347,
      "loss": 0.6123,
      "step": 4329
    },
    {
      "epoch": 1.7490530781273672,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0001954126213592233,
      "loss": 0.6836,
      "step": 4330
    },
    {
      "epoch": 1.7494570981263573,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0001953883495145631,
      "loss": 0.6794,
      "step": 4331
    },
    {
      "epoch": 1.7498611181253472,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00019536407766990288,
      "loss": 0.5777,
      "step": 4332
    },
    {
      "epoch": 1.750265138124337,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0001953398058252427,
      "loss": 0.5087,
      "step": 4333
    },
    {
      "epoch": 1.7506691581233271,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0001953155339805825,
      "loss": 0.5604,
      "step": 4334
    },
    {
      "epoch": 1.7510731781223172,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0001952912621359223,
      "loss": 0.5877,
      "step": 4335
    },
    {
      "epoch": 1.7514771981213069,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019526699029126212,
      "loss": 0.5693,
      "step": 4336
    },
    {
      "epoch": 1.751881218120297,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019524271844660192,
      "loss": 0.6514,
      "step": 4337
    },
    {
      "epoch": 1.752285238119287,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00019521844660194175,
      "loss": 0.569,
      "step": 4338
    },
    {
      "epoch": 1.7526892581182767,
      "grad_norm": 0.5625,
      "learning_rate": 0.00019519417475728155,
      "loss": 0.609,
      "step": 4339
    },
    {
      "epoch": 1.7530932781172668,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019516990291262133,
      "loss": 0.6879,
      "step": 4340
    },
    {
      "epoch": 1.7534972981162569,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00019514563106796116,
      "loss": 0.6702,
      "step": 4341
    },
    {
      "epoch": 1.7539013181152467,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019512135922330096,
      "loss": 0.6774,
      "step": 4342
    },
    {
      "epoch": 1.7543053381142366,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00019509708737864076,
      "loss": 0.6977,
      "step": 4343
    },
    {
      "epoch": 1.7547093581132267,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00019507281553398056,
      "loss": 0.6235,
      "step": 4344
    },
    {
      "epoch": 1.7551133781122166,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019504854368932036,
      "loss": 0.6604,
      "step": 4345
    },
    {
      "epoch": 1.7555173981112064,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00019502427184466017,
      "loss": 0.5341,
      "step": 4346
    },
    {
      "epoch": 1.7559214181101965,
      "grad_norm": 0.53515625,
      "learning_rate": 0.000195,
      "loss": 0.6446,
      "step": 4347
    },
    {
      "epoch": 1.7563254381091864,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00019497572815533977,
      "loss": 0.5524,
      "step": 4348
    },
    {
      "epoch": 1.7567294581081763,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0001949514563106796,
      "loss": 0.6641,
      "step": 4349
    },
    {
      "epoch": 1.7571334781071664,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0001949271844660194,
      "loss": 0.5644,
      "step": 4350
    },
    {
      "epoch": 1.7575374981061562,
      "grad_norm": 0.427734375,
      "learning_rate": 0.0001949029126213592,
      "loss": 0.5749,
      "step": 4351
    },
    {
      "epoch": 1.757941518105146,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00019487864077669904,
      "loss": 0.5015,
      "step": 4352
    },
    {
      "epoch": 1.7583455381041362,
      "grad_norm": 0.5625,
      "learning_rate": 0.0001948543689320388,
      "loss": 0.6637,
      "step": 4353
    },
    {
      "epoch": 1.758749558103126,
      "grad_norm": 0.46484375,
      "learning_rate": 0.0001948300970873786,
      "loss": 0.5448,
      "step": 4354
    },
    {
      "epoch": 1.759153578102116,
      "grad_norm": 0.46875,
      "learning_rate": 0.00019480582524271844,
      "loss": 0.5872,
      "step": 4355
    },
    {
      "epoch": 1.759557598101106,
      "grad_norm": 0.5,
      "learning_rate": 0.00019478155339805824,
      "loss": 0.5861,
      "step": 4356
    },
    {
      "epoch": 1.759961618100096,
      "grad_norm": 0.484375,
      "learning_rate": 0.00019475728155339802,
      "loss": 0.5536,
      "step": 4357
    },
    {
      "epoch": 1.7603656380990857,
      "grad_norm": 0.5625,
      "learning_rate": 0.00019473300970873785,
      "loss": 0.6411,
      "step": 4358
    },
    {
      "epoch": 1.7607696580980758,
      "grad_norm": 0.484375,
      "learning_rate": 0.00019470873786407765,
      "loss": 0.5972,
      "step": 4359
    },
    {
      "epoch": 1.761173678097066,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019468446601941748,
      "loss": 0.5727,
      "step": 4360
    },
    {
      "epoch": 1.7615776980960558,
      "grad_norm": 0.484375,
      "learning_rate": 0.00019466019417475726,
      "loss": 0.6261,
      "step": 4361
    },
    {
      "epoch": 1.7619817180950457,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00019463592233009706,
      "loss": 0.6115,
      "step": 4362
    },
    {
      "epoch": 1.7623857380940358,
      "grad_norm": 0.462890625,
      "learning_rate": 0.0001946116504854369,
      "loss": 0.4896,
      "step": 4363
    },
    {
      "epoch": 1.7627897580930256,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0001945873786407767,
      "loss": 0.6699,
      "step": 4364
    },
    {
      "epoch": 1.7631937780920155,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00019456310679611647,
      "loss": 0.5617,
      "step": 4365
    },
    {
      "epoch": 1.7635977980910056,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0001945388349514563,
      "loss": 0.572,
      "step": 4366
    },
    {
      "epoch": 1.7640018180899955,
      "grad_norm": 0.6328125,
      "learning_rate": 0.0001945145631067961,
      "loss": 0.6897,
      "step": 4367
    },
    {
      "epoch": 1.7644058380889853,
      "grad_norm": 0.53125,
      "learning_rate": 0.0001944902912621359,
      "loss": 0.6226,
      "step": 4368
    },
    {
      "epoch": 1.7648098580879754,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00019446601941747573,
      "loss": 0.5579,
      "step": 4369
    },
    {
      "epoch": 1.7652138780869653,
      "grad_norm": 0.515625,
      "learning_rate": 0.0001944417475728155,
      "loss": 0.5803,
      "step": 4370
    },
    {
      "epoch": 1.7656178980859552,
      "grad_norm": 0.59375,
      "learning_rate": 0.00019441747572815533,
      "loss": 0.6396,
      "step": 4371
    },
    {
      "epoch": 1.7660219180849452,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019439320388349514,
      "loss": 0.6675,
      "step": 4372
    },
    {
      "epoch": 1.766425938083935,
      "grad_norm": 0.5625,
      "learning_rate": 0.0001943689320388349,
      "loss": 0.663,
      "step": 4373
    },
    {
      "epoch": 1.766829958082925,
      "grad_norm": 0.515625,
      "learning_rate": 0.00019434466019417474,
      "loss": 0.5754,
      "step": 4374
    },
    {
      "epoch": 1.767233978081915,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00019432038834951454,
      "loss": 0.6463,
      "step": 4375
    },
    {
      "epoch": 1.7676379980809052,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00019429611650485435,
      "loss": 0.5425,
      "step": 4376
    },
    {
      "epoch": 1.7680420180798948,
      "grad_norm": 0.416015625,
      "learning_rate": 0.00019427184466019418,
      "loss": 0.5751,
      "step": 4377
    },
    {
      "epoch": 1.768446038078885,
      "grad_norm": 0.546875,
      "learning_rate": 0.00019424757281553395,
      "loss": 0.6292,
      "step": 4378
    },
    {
      "epoch": 1.768850058077875,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019422330097087378,
      "loss": 0.6046,
      "step": 4379
    },
    {
      "epoch": 1.7692540780768649,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019419902912621358,
      "loss": 0.6239,
      "step": 4380
    },
    {
      "epoch": 1.7696580980758547,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019417475728155338,
      "loss": 0.6019,
      "step": 4381
    },
    {
      "epoch": 1.7700621180748448,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0001941504854368932,
      "loss": 0.6274,
      "step": 4382
    },
    {
      "epoch": 1.7704661380738347,
      "grad_norm": 0.6171875,
      "learning_rate": 0.000194126213592233,
      "loss": 0.6021,
      "step": 4383
    },
    {
      "epoch": 1.7708701580728246,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0001941019417475728,
      "loss": 0.6647,
      "step": 4384
    },
    {
      "epoch": 1.7712741780718146,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019407766990291262,
      "loss": 0.5437,
      "step": 4385
    },
    {
      "epoch": 1.7716781980708045,
      "grad_norm": 0.578125,
      "learning_rate": 0.0001940533980582524,
      "loss": 0.5816,
      "step": 4386
    },
    {
      "epoch": 1.7720822180697944,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0001940291262135922,
      "loss": 0.6439,
      "step": 4387
    },
    {
      "epoch": 1.7724862380687845,
      "grad_norm": 0.578125,
      "learning_rate": 0.00019400485436893203,
      "loss": 0.582,
      "step": 4388
    },
    {
      "epoch": 1.7728902580677743,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00019398058252427183,
      "loss": 0.6177,
      "step": 4389
    },
    {
      "epoch": 1.7732942780667642,
      "grad_norm": 0.546875,
      "learning_rate": 0.00019395631067961166,
      "loss": 0.6604,
      "step": 4390
    },
    {
      "epoch": 1.7736982980657543,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00019393203883495143,
      "loss": 0.5638,
      "step": 4391
    },
    {
      "epoch": 1.7741023180647442,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00019390776699029124,
      "loss": 0.5843,
      "step": 4392
    },
    {
      "epoch": 1.774506338063734,
      "grad_norm": 0.451171875,
      "learning_rate": 0.00019388349514563107,
      "loss": 0.5678,
      "step": 4393
    },
    {
      "epoch": 1.7749103580627241,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00019385922330097087,
      "loss": 0.5805,
      "step": 4394
    },
    {
      "epoch": 1.7753143780617142,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019383495145631064,
      "loss": 0.6078,
      "step": 4395
    },
    {
      "epoch": 1.7757183980607039,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00019381067961165047,
      "loss": 0.6096,
      "step": 4396
    },
    {
      "epoch": 1.776122418059694,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00019378640776699028,
      "loss": 0.6692,
      "step": 4397
    },
    {
      "epoch": 1.776526438058684,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00019376213592233005,
      "loss": 0.6724,
      "step": 4398
    },
    {
      "epoch": 1.7769304580576737,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00019373786407766988,
      "loss": 0.5543,
      "step": 4399
    },
    {
      "epoch": 1.7773344780566638,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00019371359223300968,
      "loss": 0.703,
      "step": 4400
    },
    {
      "epoch": 1.7777384980556539,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0001936893203883495,
      "loss": 0.5532,
      "step": 4401
    },
    {
      "epoch": 1.7781425180546437,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019366504854368931,
      "loss": 0.6188,
      "step": 4402
    },
    {
      "epoch": 1.7785465380536336,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0001936407766990291,
      "loss": 0.5519,
      "step": 4403
    },
    {
      "epoch": 1.7789505580526237,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00019361650485436892,
      "loss": 0.6434,
      "step": 4404
    },
    {
      "epoch": 1.7793545780516136,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00019359223300970872,
      "loss": 0.6295,
      "step": 4405
    },
    {
      "epoch": 1.7797585980506034,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00019356796116504852,
      "loss": 0.7072,
      "step": 4406
    },
    {
      "epoch": 1.7801626180495935,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00019354368932038833,
      "loss": 0.5749,
      "step": 4407
    },
    {
      "epoch": 1.7805666380485834,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00019351941747572813,
      "loss": 0.6727,
      "step": 4408
    },
    {
      "epoch": 1.7809706580475733,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00019349514563106796,
      "loss": 0.6101,
      "step": 4409
    },
    {
      "epoch": 1.7813746780465634,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00019347087378640776,
      "loss": 0.6239,
      "step": 4410
    },
    {
      "epoch": 1.7817786980455532,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00019344660194174754,
      "loss": 0.5985,
      "step": 4411
    },
    {
      "epoch": 1.782182718044543,
      "grad_norm": 0.484375,
      "learning_rate": 0.00019342233009708737,
      "loss": 0.6285,
      "step": 4412
    },
    {
      "epoch": 1.7825867380435332,
      "grad_norm": 0.43359375,
      "learning_rate": 0.00019339805825242717,
      "loss": 0.5521,
      "step": 4413
    },
    {
      "epoch": 1.782990758042523,
      "grad_norm": 0.546875,
      "learning_rate": 0.00019337378640776697,
      "loss": 0.6153,
      "step": 4414
    },
    {
      "epoch": 1.783394778041513,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0001933495145631068,
      "loss": 0.6497,
      "step": 4415
    },
    {
      "epoch": 1.783798798040503,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00019332524271844657,
      "loss": 0.5736,
      "step": 4416
    },
    {
      "epoch": 1.784202818039493,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00019330097087378638,
      "loss": 0.6196,
      "step": 4417
    },
    {
      "epoch": 1.7846068380384827,
      "grad_norm": 0.671875,
      "learning_rate": 0.0001932766990291262,
      "loss": 0.7193,
      "step": 4418
    },
    {
      "epoch": 1.7850108580374728,
      "grad_norm": 0.46875,
      "learning_rate": 0.000193252427184466,
      "loss": 0.5981,
      "step": 4419
    },
    {
      "epoch": 1.785414878036463,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0001932281553398058,
      "loss": 0.6164,
      "step": 4420
    },
    {
      "epoch": 1.7858188980354528,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00019320388349514561,
      "loss": 0.6517,
      "step": 4421
    },
    {
      "epoch": 1.7862229180344427,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00019317961165048542,
      "loss": 0.6348,
      "step": 4422
    },
    {
      "epoch": 1.7866269380334328,
      "grad_norm": 0.435546875,
      "learning_rate": 0.00019315533980582525,
      "loss": 0.6268,
      "step": 4423
    },
    {
      "epoch": 1.7870309580324226,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00019313106796116502,
      "loss": 0.6205,
      "step": 4424
    },
    {
      "epoch": 1.7874349780314125,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00019310679611650482,
      "loss": 0.624,
      "step": 4425
    },
    {
      "epoch": 1.7878389980304026,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019308252427184465,
      "loss": 0.6648,
      "step": 4426
    },
    {
      "epoch": 1.7882430180293925,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00019305825242718445,
      "loss": 0.597,
      "step": 4427
    },
    {
      "epoch": 1.7886470380283823,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00019303398058252423,
      "loss": 0.6712,
      "step": 4428
    },
    {
      "epoch": 1.7890510580273724,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00019300970873786406,
      "loss": 0.6216,
      "step": 4429
    },
    {
      "epoch": 1.7894550780263623,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00019298543689320386,
      "loss": 0.6468,
      "step": 4430
    },
    {
      "epoch": 1.7898590980253521,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0001929611650485437,
      "loss": 0.6169,
      "step": 4431
    },
    {
      "epoch": 1.7902631180243422,
      "grad_norm": 0.478515625,
      "learning_rate": 0.0001929368932038835,
      "loss": 0.5881,
      "step": 4432
    },
    {
      "epoch": 1.790667138023332,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019291262135922327,
      "loss": 0.5926,
      "step": 4433
    },
    {
      "epoch": 1.791071158022322,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0001928883495145631,
      "loss": 0.6746,
      "step": 4434
    },
    {
      "epoch": 1.791475178021312,
      "grad_norm": 0.46484375,
      "learning_rate": 0.0001928640776699029,
      "loss": 0.5416,
      "step": 4435
    },
    {
      "epoch": 1.7918791980203022,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00019283980582524268,
      "loss": 0.5604,
      "step": 4436
    },
    {
      "epoch": 1.7922832180192918,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0001928155339805825,
      "loss": 0.5756,
      "step": 4437
    },
    {
      "epoch": 1.792687238018282,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0001927912621359223,
      "loss": 0.6492,
      "step": 4438
    },
    {
      "epoch": 1.793091258017272,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00019276699029126214,
      "loss": 0.5834,
      "step": 4439
    },
    {
      "epoch": 1.7934952780162619,
      "grad_norm": 0.5,
      "learning_rate": 0.00019274271844660194,
      "loss": 0.512,
      "step": 4440
    },
    {
      "epoch": 1.7938992980152517,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00019271844660194171,
      "loss": 0.5548,
      "step": 4441
    },
    {
      "epoch": 1.7943033180142418,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00019269417475728154,
      "loss": 0.6352,
      "step": 4442
    },
    {
      "epoch": 1.7947073380132317,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00019266990291262135,
      "loss": 0.6609,
      "step": 4443
    },
    {
      "epoch": 1.7951113580122215,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00019264563106796115,
      "loss": 0.6094,
      "step": 4444
    },
    {
      "epoch": 1.7955153780112116,
      "grad_norm": 0.63671875,
      "learning_rate": 0.00019262135922330095,
      "loss": 0.69,
      "step": 4445
    },
    {
      "epoch": 1.7959193980102015,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00019259708737864075,
      "loss": 0.6269,
      "step": 4446
    },
    {
      "epoch": 1.7963234180091914,
      "grad_norm": 0.578125,
      "learning_rate": 0.00019257281553398056,
      "loss": 0.6288,
      "step": 4447
    },
    {
      "epoch": 1.7967274380081815,
      "grad_norm": 0.388671875,
      "learning_rate": 0.00019254854368932039,
      "loss": 0.5247,
      "step": 4448
    },
    {
      "epoch": 1.7971314580071713,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00019252427184466016,
      "loss": 0.6968,
      "step": 4449
    },
    {
      "epoch": 1.7975354780061612,
      "grad_norm": 0.51953125,
      "learning_rate": 0.0001925,
      "loss": 0.641,
      "step": 4450
    },
    {
      "epoch": 1.7979394980051513,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0001924757281553398,
      "loss": 0.5877,
      "step": 4451
    },
    {
      "epoch": 1.7983435180041412,
      "grad_norm": 0.47265625,
      "learning_rate": 0.0001924514563106796,
      "loss": 0.5945,
      "step": 4452
    },
    {
      "epoch": 1.798747538003131,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00019242718446601942,
      "loss": 0.5583,
      "step": 4453
    },
    {
      "epoch": 1.7991515580021211,
      "grad_norm": 0.703125,
      "learning_rate": 0.0001924029126213592,
      "loss": 0.72,
      "step": 4454
    },
    {
      "epoch": 1.7995555780011112,
      "grad_norm": 0.69140625,
      "learning_rate": 0.000192378640776699,
      "loss": 0.695,
      "step": 4455
    },
    {
      "epoch": 1.7999595980001009,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019235436893203883,
      "loss": 0.6164,
      "step": 4456
    },
    {
      "epoch": 1.800363617999091,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00019233009708737863,
      "loss": 0.6614,
      "step": 4457
    },
    {
      "epoch": 1.800767637998081,
      "grad_norm": 0.458984375,
      "learning_rate": 0.0001923058252427184,
      "loss": 0.6055,
      "step": 4458
    },
    {
      "epoch": 1.801171657997071,
      "grad_norm": 0.578125,
      "learning_rate": 0.00019228155339805824,
      "loss": 0.6586,
      "step": 4459
    },
    {
      "epoch": 1.8015756779960608,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00019225728155339804,
      "loss": 0.6417,
      "step": 4460
    },
    {
      "epoch": 1.8019796979950509,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00019223300970873787,
      "loss": 0.6007,
      "step": 4461
    },
    {
      "epoch": 1.8023837179940407,
      "grad_norm": 0.546875,
      "learning_rate": 0.00019220873786407765,
      "loss": 0.5834,
      "step": 4462
    },
    {
      "epoch": 1.8027877379930306,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00019218446601941745,
      "loss": 0.6595,
      "step": 4463
    },
    {
      "epoch": 1.8031917579920207,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00019216019417475728,
      "loss": 0.5303,
      "step": 4464
    },
    {
      "epoch": 1.8035957779910106,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019213592233009708,
      "loss": 0.6588,
      "step": 4465
    },
    {
      "epoch": 1.8039997979900004,
      "grad_norm": 0.41015625,
      "learning_rate": 0.00019211165048543685,
      "loss": 0.482,
      "step": 4466
    },
    {
      "epoch": 1.8044038179889905,
      "grad_norm": 0.5,
      "learning_rate": 0.00019208737864077668,
      "loss": 0.5017,
      "step": 4467
    },
    {
      "epoch": 1.8048078379879804,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019206310679611649,
      "loss": 0.6329,
      "step": 4468
    },
    {
      "epoch": 1.8052118579869703,
      "grad_norm": 0.546875,
      "learning_rate": 0.0001920388349514563,
      "loss": 0.6062,
      "step": 4469
    },
    {
      "epoch": 1.8056158779859603,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0001920145631067961,
      "loss": 0.6541,
      "step": 4470
    },
    {
      "epoch": 1.8060198979849502,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0001919902912621359,
      "loss": 0.6053,
      "step": 4471
    },
    {
      "epoch": 1.80642391798394,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00019196601941747572,
      "loss": 0.6753,
      "step": 4472
    },
    {
      "epoch": 1.8068279379829302,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00019194174757281553,
      "loss": 0.6436,
      "step": 4473
    },
    {
      "epoch": 1.8072319579819203,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0001919174757281553,
      "loss": 0.6639,
      "step": 4474
    },
    {
      "epoch": 1.80763597798091,
      "grad_norm": 0.484375,
      "learning_rate": 0.00019189320388349513,
      "loss": 0.6331,
      "step": 4475
    },
    {
      "epoch": 1.8080399979799,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00019186893203883493,
      "loss": 0.6678,
      "step": 4476
    },
    {
      "epoch": 1.80844401797889,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00019184466019417473,
      "loss": 0.6958,
      "step": 4477
    },
    {
      "epoch": 1.8088480379778797,
      "grad_norm": 0.484375,
      "learning_rate": 0.00019182038834951456,
      "loss": 0.6148,
      "step": 4478
    },
    {
      "epoch": 1.8092520579768698,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00019179611650485434,
      "loss": 0.6177,
      "step": 4479
    },
    {
      "epoch": 1.80965607797586,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00019177184466019417,
      "loss": 0.6018,
      "step": 4480
    },
    {
      "epoch": 1.8100600979748498,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00019174757281553397,
      "loss": 0.6429,
      "step": 4481
    },
    {
      "epoch": 1.8104641179738397,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019172330097087377,
      "loss": 0.6173,
      "step": 4482
    },
    {
      "epoch": 1.8108681379728298,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00019169902912621358,
      "loss": 0.6407,
      "step": 4483
    },
    {
      "epoch": 1.8112721579718196,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00019167475728155338,
      "loss": 0.6209,
      "step": 4484
    },
    {
      "epoch": 1.8116761779708095,
      "grad_norm": 0.5625,
      "learning_rate": 0.00019165048543689318,
      "loss": 0.6814,
      "step": 4485
    },
    {
      "epoch": 1.8120801979697996,
      "grad_norm": 0.50390625,
      "learning_rate": 0.000191626213592233,
      "loss": 0.5928,
      "step": 4486
    },
    {
      "epoch": 1.8124842179687894,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00019160194174757278,
      "loss": 0.65,
      "step": 4487
    },
    {
      "epoch": 1.8128882379677793,
      "grad_norm": 0.53125,
      "learning_rate": 0.0001915776699029126,
      "loss": 0.6111,
      "step": 4488
    },
    {
      "epoch": 1.8132922579667694,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00019155339805825242,
      "loss": 0.6041,
      "step": 4489
    },
    {
      "epoch": 1.8136962779657593,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019152912621359222,
      "loss": 0.6118,
      "step": 4490
    },
    {
      "epoch": 1.8141002979647491,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00019150485436893205,
      "loss": 0.637,
      "step": 4491
    },
    {
      "epoch": 1.8145043179637392,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00019148058252427182,
      "loss": 0.561,
      "step": 4492
    },
    {
      "epoch": 1.814908337962729,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00019145631067961163,
      "loss": 0.6401,
      "step": 4493
    },
    {
      "epoch": 1.815312357961719,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00019143203883495146,
      "loss": 0.6385,
      "step": 4494
    },
    {
      "epoch": 1.815716377960709,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00019140776699029126,
      "loss": 0.553,
      "step": 4495
    },
    {
      "epoch": 1.8161203979596992,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00019138349514563103,
      "loss": 0.6017,
      "step": 4496
    },
    {
      "epoch": 1.8165244179586888,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00019135922330097086,
      "loss": 0.6372,
      "step": 4497
    },
    {
      "epoch": 1.8169284379576789,
      "grad_norm": 0.484375,
      "learning_rate": 0.00019133495145631066,
      "loss": 0.5955,
      "step": 4498
    },
    {
      "epoch": 1.817332457956669,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00019131067961165044,
      "loss": 0.6753,
      "step": 4499
    },
    {
      "epoch": 1.8177364779556588,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00019128640776699027,
      "loss": 0.5713,
      "step": 4500
    },
    {
      "epoch": 1.8181404979546487,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019126213592233007,
      "loss": 0.5865,
      "step": 4501
    },
    {
      "epoch": 1.8185445179536388,
      "grad_norm": 0.478515625,
      "learning_rate": 0.0001912378640776699,
      "loss": 0.5909,
      "step": 4502
    },
    {
      "epoch": 1.8189485379526287,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0001912135922330097,
      "loss": 0.6465,
      "step": 4503
    },
    {
      "epoch": 1.8193525579516185,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00019118932038834948,
      "loss": 0.6115,
      "step": 4504
    },
    {
      "epoch": 1.8197565779506086,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0001911650485436893,
      "loss": 0.6047,
      "step": 4505
    },
    {
      "epoch": 1.8201605979495985,
      "grad_norm": 0.5,
      "learning_rate": 0.0001911407766990291,
      "loss": 0.6209,
      "step": 4506
    },
    {
      "epoch": 1.8205646179485884,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0001911165048543689,
      "loss": 0.5896,
      "step": 4507
    },
    {
      "epoch": 1.8209686379475785,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00019109223300970872,
      "loss": 0.6246,
      "step": 4508
    },
    {
      "epoch": 1.8213726579465683,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019106796116504852,
      "loss": 0.7146,
      "step": 4509
    },
    {
      "epoch": 1.8217766779455582,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00019104368932038835,
      "loss": 0.6379,
      "step": 4510
    },
    {
      "epoch": 1.8221806979445483,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00019101941747572815,
      "loss": 0.6219,
      "step": 4511
    },
    {
      "epoch": 1.8225847179435382,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00019099514563106792,
      "loss": 0.6937,
      "step": 4512
    },
    {
      "epoch": 1.822988737942528,
      "grad_norm": 0.453125,
      "learning_rate": 0.00019097087378640775,
      "loss": 0.5628,
      "step": 4513
    },
    {
      "epoch": 1.8233927579415181,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00019094660194174756,
      "loss": 0.5738,
      "step": 4514
    },
    {
      "epoch": 1.8237967779405082,
      "grad_norm": 0.59375,
      "learning_rate": 0.00019092233009708736,
      "loss": 0.697,
      "step": 4515
    },
    {
      "epoch": 1.8242007979394979,
      "grad_norm": 0.44921875,
      "learning_rate": 0.0001908980582524272,
      "loss": 0.585,
      "step": 4516
    },
    {
      "epoch": 1.824604817938488,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00019087378640776696,
      "loss": 0.6062,
      "step": 4517
    },
    {
      "epoch": 1.825008837937478,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019084951456310677,
      "loss": 0.6204,
      "step": 4518
    },
    {
      "epoch": 1.825412857936468,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0001908252427184466,
      "loss": 0.691,
      "step": 4519
    },
    {
      "epoch": 1.8258168779354578,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0001908009708737864,
      "loss": 0.6391,
      "step": 4520
    },
    {
      "epoch": 1.8262208979344479,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0001907766990291262,
      "loss": 0.6266,
      "step": 4521
    },
    {
      "epoch": 1.8266249179334377,
      "grad_norm": 0.5,
      "learning_rate": 0.000190752427184466,
      "loss": 0.64,
      "step": 4522
    },
    {
      "epoch": 1.8270289379324276,
      "grad_norm": 0.484375,
      "learning_rate": 0.0001907281553398058,
      "loss": 0.6004,
      "step": 4523
    },
    {
      "epoch": 1.8274329579314177,
      "grad_norm": 0.53125,
      "learning_rate": 0.00019070388349514563,
      "loss": 0.6428,
      "step": 4524
    },
    {
      "epoch": 1.8278369779304076,
      "grad_norm": 0.5859375,
      "learning_rate": 0.0001906796116504854,
      "loss": 0.6656,
      "step": 4525
    },
    {
      "epoch": 1.8282409979293974,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0001906553398058252,
      "loss": 0.5566,
      "step": 4526
    },
    {
      "epoch": 1.8286450179283875,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019063106796116504,
      "loss": 0.6297,
      "step": 4527
    },
    {
      "epoch": 1.8290490379273774,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00019060679611650484,
      "loss": 0.6324,
      "step": 4528
    },
    {
      "epoch": 1.8294530579263673,
      "grad_norm": 0.625,
      "learning_rate": 0.00019058252427184462,
      "loss": 0.6555,
      "step": 4529
    },
    {
      "epoch": 1.8298570779253573,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019055825242718445,
      "loss": 0.6293,
      "step": 4530
    },
    {
      "epoch": 1.8302610979243472,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00019053398058252425,
      "loss": 0.6753,
      "step": 4531
    },
    {
      "epoch": 1.830665117923337,
      "grad_norm": 0.71875,
      "learning_rate": 0.00019050970873786408,
      "loss": 0.6713,
      "step": 4532
    },
    {
      "epoch": 1.8310691379223272,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019048543689320388,
      "loss": 0.5463,
      "step": 4533
    },
    {
      "epoch": 1.8314731579213173,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00019046116504854366,
      "loss": 0.6034,
      "step": 4534
    },
    {
      "epoch": 1.831877177920307,
      "grad_norm": 0.4296875,
      "learning_rate": 0.0001904368932038835,
      "loss": 0.5397,
      "step": 4535
    },
    {
      "epoch": 1.832281197919297,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0001904126213592233,
      "loss": 0.6753,
      "step": 4536
    },
    {
      "epoch": 1.832685217918287,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00019038834951456306,
      "loss": 0.5854,
      "step": 4537
    },
    {
      "epoch": 1.833089237917277,
      "grad_norm": 0.494140625,
      "learning_rate": 0.0001903640776699029,
      "loss": 0.6046,
      "step": 4538
    },
    {
      "epoch": 1.8334932579162668,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0001903398058252427,
      "loss": 0.6685,
      "step": 4539
    },
    {
      "epoch": 1.833897277915257,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00019031553398058253,
      "loss": 0.6471,
      "step": 4540
    },
    {
      "epoch": 1.8343012979142468,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00019029126213592233,
      "loss": 0.6748,
      "step": 4541
    },
    {
      "epoch": 1.8347053179132367,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0001902669902912621,
      "loss": 0.6187,
      "step": 4542
    },
    {
      "epoch": 1.8351093379122267,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00019024271844660193,
      "loss": 0.6376,
      "step": 4543
    },
    {
      "epoch": 1.8355133579112166,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00019021844660194174,
      "loss": 0.5722,
      "step": 4544
    },
    {
      "epoch": 1.8359173779102065,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00019019417475728154,
      "loss": 0.5946,
      "step": 4545
    },
    {
      "epoch": 1.8363213979091966,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00019016990291262134,
      "loss": 0.649,
      "step": 4546
    },
    {
      "epoch": 1.8367254179081864,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00019014563106796114,
      "loss": 0.6391,
      "step": 4547
    },
    {
      "epoch": 1.8371294379071763,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00019012135922330094,
      "loss": 0.6373,
      "step": 4548
    },
    {
      "epoch": 1.8375334579061664,
      "grad_norm": 0.435546875,
      "learning_rate": 0.00019009708737864077,
      "loss": 0.5726,
      "step": 4549
    },
    {
      "epoch": 1.8379374779051563,
      "grad_norm": 0.53125,
      "learning_rate": 0.00019007281553398055,
      "loss": 0.5689,
      "step": 4550
    },
    {
      "epoch": 1.8383414979041461,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00019004854368932038,
      "loss": 0.6914,
      "step": 4551
    },
    {
      "epoch": 1.8387455179031362,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00019002427184466018,
      "loss": 0.6539,
      "step": 4552
    },
    {
      "epoch": 1.8391495379021263,
      "grad_norm": 0.53125,
      "learning_rate": 0.00018999999999999998,
      "loss": 0.5913,
      "step": 4553
    },
    {
      "epoch": 1.839553557901116,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0001899757281553398,
      "loss": 0.6389,
      "step": 4554
    },
    {
      "epoch": 1.839957577900106,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0001899514563106796,
      "loss": 0.6279,
      "step": 4555
    },
    {
      "epoch": 1.8403615978990961,
      "grad_norm": 0.478515625,
      "learning_rate": 0.0001899271844660194,
      "loss": 0.5743,
      "step": 4556
    },
    {
      "epoch": 1.8407656178980858,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00018990291262135922,
      "loss": 0.5965,
      "step": 4557
    },
    {
      "epoch": 1.8411696378970759,
      "grad_norm": 0.53125,
      "learning_rate": 0.00018987864077669902,
      "loss": 0.6032,
      "step": 4558
    },
    {
      "epoch": 1.841573657896066,
      "grad_norm": 0.6875,
      "learning_rate": 0.0001898543689320388,
      "loss": 0.6457,
      "step": 4559
    },
    {
      "epoch": 1.8419776778950558,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00018983009708737863,
      "loss": 0.5781,
      "step": 4560
    },
    {
      "epoch": 1.8423816978940457,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00018980582524271843,
      "loss": 0.5983,
      "step": 4561
    },
    {
      "epoch": 1.8427857178930358,
      "grad_norm": 0.439453125,
      "learning_rate": 0.00018978155339805826,
      "loss": 0.5904,
      "step": 4562
    },
    {
      "epoch": 1.8431897378920257,
      "grad_norm": 0.439453125,
      "learning_rate": 0.00018975728155339803,
      "loss": 0.6336,
      "step": 4563
    },
    {
      "epoch": 1.8435937578910155,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018973300970873784,
      "loss": 0.6736,
      "step": 4564
    },
    {
      "epoch": 1.8439977778900056,
      "grad_norm": 0.484375,
      "learning_rate": 0.00018970873786407767,
      "loss": 0.5521,
      "step": 4565
    },
    {
      "epoch": 1.8444017978889955,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018968446601941747,
      "loss": 0.6379,
      "step": 4566
    },
    {
      "epoch": 1.8448058178879854,
      "grad_norm": 0.546875,
      "learning_rate": 0.00018966019417475724,
      "loss": 0.6871,
      "step": 4567
    },
    {
      "epoch": 1.8452098378869755,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00018963592233009707,
      "loss": 0.5311,
      "step": 4568
    },
    {
      "epoch": 1.8456138578859653,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00018961165048543688,
      "loss": 0.5928,
      "step": 4569
    },
    {
      "epoch": 1.8460178778849552,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00018958737864077668,
      "loss": 0.6423,
      "step": 4570
    },
    {
      "epoch": 1.8464218978839453,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00018956310679611648,
      "loss": 0.6155,
      "step": 4571
    },
    {
      "epoch": 1.8468259178829352,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00018953883495145628,
      "loss": 0.642,
      "step": 4572
    },
    {
      "epoch": 1.847229937881925,
      "grad_norm": 0.53125,
      "learning_rate": 0.0001895145631067961,
      "loss": 0.609,
      "step": 4573
    },
    {
      "epoch": 1.8476339578809151,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00018949029126213591,
      "loss": 0.5377,
      "step": 4574
    },
    {
      "epoch": 1.8480379778799052,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0001894660194174757,
      "loss": 0.5993,
      "step": 4575
    },
    {
      "epoch": 1.8484419978788948,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00018944174757281552,
      "loss": 0.6855,
      "step": 4576
    },
    {
      "epoch": 1.848846017877885,
      "grad_norm": 0.443359375,
      "learning_rate": 0.00018941747572815532,
      "loss": 0.5707,
      "step": 4577
    },
    {
      "epoch": 1.849250037876875,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00018939320388349512,
      "loss": 0.5973,
      "step": 4578
    },
    {
      "epoch": 1.849654057875865,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00018936893203883495,
      "loss": 0.6816,
      "step": 4579
    },
    {
      "epoch": 1.8500580778748548,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00018934466019417473,
      "loss": 0.5946,
      "step": 4580
    },
    {
      "epoch": 1.8504620978738449,
      "grad_norm": 0.416015625,
      "learning_rate": 0.00018932038834951456,
      "loss": 0.5885,
      "step": 4581
    },
    {
      "epoch": 1.8508661178728347,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00018929611650485436,
      "loss": 0.6183,
      "step": 4582
    },
    {
      "epoch": 1.8512701378718246,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00018927184466019416,
      "loss": 0.7045,
      "step": 4583
    },
    {
      "epoch": 1.8516741578708147,
      "grad_norm": 0.431640625,
      "learning_rate": 0.00018924757281553396,
      "loss": 0.6132,
      "step": 4584
    },
    {
      "epoch": 1.8520781778698046,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00018922330097087377,
      "loss": 0.6642,
      "step": 4585
    },
    {
      "epoch": 1.8524821978687944,
      "grad_norm": 0.44140625,
      "learning_rate": 0.00018919902912621357,
      "loss": 0.5905,
      "step": 4586
    },
    {
      "epoch": 1.8528862178677845,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0001891747572815534,
      "loss": 0.6681,
      "step": 4587
    },
    {
      "epoch": 1.8532902378667744,
      "grad_norm": 0.453125,
      "learning_rate": 0.00018915048543689317,
      "loss": 0.5879,
      "step": 4588
    },
    {
      "epoch": 1.8536942578657642,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00018912621359223298,
      "loss": 0.5853,
      "step": 4589
    },
    {
      "epoch": 1.8540982778647543,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0001891019417475728,
      "loss": 0.6134,
      "step": 4590
    },
    {
      "epoch": 1.8545022978637442,
      "grad_norm": 0.53515625,
      "learning_rate": 0.0001890776699029126,
      "loss": 0.6569,
      "step": 4591
    },
    {
      "epoch": 1.854906317862734,
      "grad_norm": 35.25,
      "learning_rate": 0.00018905339805825244,
      "loss": 0.5831,
      "step": 4592
    },
    {
      "epoch": 1.8553103378617242,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0001890291262135922,
      "loss": 0.5924,
      "step": 4593
    },
    {
      "epoch": 1.8557143578607143,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018900485436893201,
      "loss": 0.5976,
      "step": 4594
    },
    {
      "epoch": 1.856118377859704,
      "grad_norm": 0.5625,
      "learning_rate": 0.00018898058252427184,
      "loss": 0.5996,
      "step": 4595
    },
    {
      "epoch": 1.856522397858694,
      "grad_norm": 0.5625,
      "learning_rate": 0.00018895631067961165,
      "loss": 0.6146,
      "step": 4596
    },
    {
      "epoch": 1.856926417857684,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00018893203883495142,
      "loss": 0.5797,
      "step": 4597
    },
    {
      "epoch": 1.857330437856674,
      "grad_norm": 0.671875,
      "learning_rate": 0.00018890776699029125,
      "loss": 0.6574,
      "step": 4598
    },
    {
      "epoch": 1.8577344578556638,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00018888349514563105,
      "loss": 0.6288,
      "step": 4599
    },
    {
      "epoch": 1.858138477854654,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00018885922330097083,
      "loss": 0.6412,
      "step": 4600
    },
    {
      "epoch": 1.8585424978536438,
      "grad_norm": 0.427734375,
      "learning_rate": 0.00018883495145631066,
      "loss": 0.6159,
      "step": 4601
    },
    {
      "epoch": 1.8589465178526337,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00018881067961165046,
      "loss": 0.6143,
      "step": 4602
    },
    {
      "epoch": 1.8593505378516237,
      "grad_norm": 0.498046875,
      "learning_rate": 0.0001887864077669903,
      "loss": 0.626,
      "step": 4603
    },
    {
      "epoch": 1.8597545578506136,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0001887621359223301,
      "loss": 0.6385,
      "step": 4604
    },
    {
      "epoch": 1.8601585778496035,
      "grad_norm": 0.45703125,
      "learning_rate": 0.00018873786407766987,
      "loss": 0.5892,
      "step": 4605
    },
    {
      "epoch": 1.8605625978485936,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0001887135922330097,
      "loss": 0.6808,
      "step": 4606
    },
    {
      "epoch": 1.8609666178475834,
      "grad_norm": 0.46875,
      "learning_rate": 0.0001886893203883495,
      "loss": 0.5921,
      "step": 4607
    },
    {
      "epoch": 1.8613706378465733,
      "grad_norm": 0.4453125,
      "learning_rate": 0.0001886650485436893,
      "loss": 0.5776,
      "step": 4608
    },
    {
      "epoch": 1.8617746578455634,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0001886407766990291,
      "loss": 0.6412,
      "step": 4609
    },
    {
      "epoch": 1.8621786778445533,
      "grad_norm": 0.515625,
      "learning_rate": 0.0001886165048543689,
      "loss": 0.5503,
      "step": 4610
    },
    {
      "epoch": 1.8625826978435431,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00018859223300970874,
      "loss": 0.6488,
      "step": 4611
    },
    {
      "epoch": 1.8629867178425332,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018856796116504854,
      "loss": 0.6673,
      "step": 4612
    },
    {
      "epoch": 1.8633907378415233,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0001885436893203883,
      "loss": 0.5831,
      "step": 4613
    },
    {
      "epoch": 1.863794757840513,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00018851941747572814,
      "loss": 0.5515,
      "step": 4614
    },
    {
      "epoch": 1.864198777839503,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018849514563106795,
      "loss": 0.53,
      "step": 4615
    },
    {
      "epoch": 1.8646027978384931,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00018847087378640775,
      "loss": 0.5976,
      "step": 4616
    },
    {
      "epoch": 1.865006817837483,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00018844660194174758,
      "loss": 0.6714,
      "step": 4617
    },
    {
      "epoch": 1.8654108378364729,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00018842233009708735,
      "loss": 0.5284,
      "step": 4618
    },
    {
      "epoch": 1.865814857835463,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00018839805825242715,
      "loss": 0.5952,
      "step": 4619
    },
    {
      "epoch": 1.8662188778344528,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00018837378640776698,
      "loss": 0.6395,
      "step": 4620
    },
    {
      "epoch": 1.8666228978334427,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0001883495145631068,
      "loss": 0.5695,
      "step": 4621
    },
    {
      "epoch": 1.8670269178324328,
      "grad_norm": 0.484375,
      "learning_rate": 0.0001883252427184466,
      "loss": 0.5747,
      "step": 4622
    },
    {
      "epoch": 1.8674309378314227,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0001883009708737864,
      "loss": 0.5903,
      "step": 4623
    },
    {
      "epoch": 1.8678349578304125,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0001882766990291262,
      "loss": 0.5895,
      "step": 4624
    },
    {
      "epoch": 1.8682389778294026,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00018825242718446602,
      "loss": 0.6891,
      "step": 4625
    },
    {
      "epoch": 1.8686429978283925,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0001882281553398058,
      "loss": 0.6493,
      "step": 4626
    },
    {
      "epoch": 1.8690470178273824,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0001882038834951456,
      "loss": 0.6532,
      "step": 4627
    },
    {
      "epoch": 1.8694510378263725,
      "grad_norm": 0.4765625,
      "learning_rate": 0.00018817961165048543,
      "loss": 0.6218,
      "step": 4628
    },
    {
      "epoch": 1.8698550578253623,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00018815533980582523,
      "loss": 0.5613,
      "step": 4629
    },
    {
      "epoch": 1.8702590778243522,
      "grad_norm": 0.431640625,
      "learning_rate": 0.000188131067961165,
      "loss": 0.5554,
      "step": 4630
    },
    {
      "epoch": 1.8706630978233423,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018810679611650484,
      "loss": 0.663,
      "step": 4631
    },
    {
      "epoch": 1.8710671178223324,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00018808252427184464,
      "loss": 0.6132,
      "step": 4632
    },
    {
      "epoch": 1.871471137821322,
      "grad_norm": 0.423828125,
      "learning_rate": 0.00018805825242718447,
      "loss": 0.567,
      "step": 4633
    },
    {
      "epoch": 1.871875157820312,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00018803398058252424,
      "loss": 0.5978,
      "step": 4634
    },
    {
      "epoch": 1.8722791778193022,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00018800970873786405,
      "loss": 0.6488,
      "step": 4635
    },
    {
      "epoch": 1.8726831978182918,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00018798543689320388,
      "loss": 0.6097,
      "step": 4636
    },
    {
      "epoch": 1.873087217817282,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00018796116504854368,
      "loss": 0.6448,
      "step": 4637
    },
    {
      "epoch": 1.873491237816272,
      "grad_norm": 0.546875,
      "learning_rate": 0.00018793689320388345,
      "loss": 0.6652,
      "step": 4638
    },
    {
      "epoch": 1.873895257815262,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00018791262135922328,
      "loss": 0.6804,
      "step": 4639
    },
    {
      "epoch": 1.8742992778142518,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018788834951456309,
      "loss": 0.6402,
      "step": 4640
    },
    {
      "epoch": 1.8747032978132419,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00018786407766990291,
      "loss": 0.6747,
      "step": 4641
    },
    {
      "epoch": 1.8751073178122317,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00018783980582524272,
      "loss": 0.5224,
      "step": 4642
    },
    {
      "epoch": 1.8755113378112216,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0001878155339805825,
      "loss": 0.7081,
      "step": 4643
    },
    {
      "epoch": 1.8759153578102117,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00018779126213592232,
      "loss": 0.6241,
      "step": 4644
    },
    {
      "epoch": 1.8763193778092015,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00018776699029126212,
      "loss": 0.607,
      "step": 4645
    },
    {
      "epoch": 1.8767233978081914,
      "grad_norm": 0.435546875,
      "learning_rate": 0.00018774271844660193,
      "loss": 0.6235,
      "step": 4646
    },
    {
      "epoch": 1.8771274178071815,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018771844660194173,
      "loss": 0.6208,
      "step": 4647
    },
    {
      "epoch": 1.8775314378061714,
      "grad_norm": 0.48046875,
      "learning_rate": 0.00018769417475728153,
      "loss": 0.5288,
      "step": 4648
    },
    {
      "epoch": 1.8779354578051612,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00018766990291262133,
      "loss": 0.6275,
      "step": 4649
    },
    {
      "epoch": 1.8783394778041513,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00018764563106796116,
      "loss": 0.585,
      "step": 4650
    },
    {
      "epoch": 1.8787434978031412,
      "grad_norm": 0.578125,
      "learning_rate": 0.00018762135922330094,
      "loss": 0.6215,
      "step": 4651
    },
    {
      "epoch": 1.879147517802131,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00018759708737864077,
      "loss": 0.6432,
      "step": 4652
    },
    {
      "epoch": 1.8795515378011212,
      "grad_norm": 0.494140625,
      "learning_rate": 0.00018757281553398057,
      "loss": 0.622,
      "step": 4653
    },
    {
      "epoch": 1.8799555578001113,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00018754854368932037,
      "loss": 0.605,
      "step": 4654
    },
    {
      "epoch": 1.880359577799101,
      "grad_norm": 0.43359375,
      "learning_rate": 0.0001875242718446602,
      "loss": 0.5662,
      "step": 4655
    },
    {
      "epoch": 1.880763597798091,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018749999999999998,
      "loss": 0.6474,
      "step": 4656
    },
    {
      "epoch": 1.881167617797081,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00018747572815533978,
      "loss": 0.6551,
      "step": 4657
    },
    {
      "epoch": 1.881571637796071,
      "grad_norm": 0.515625,
      "learning_rate": 0.0001874514563106796,
      "loss": 0.6211,
      "step": 4658
    },
    {
      "epoch": 1.8819756577950608,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0001874271844660194,
      "loss": 0.5792,
      "step": 4659
    },
    {
      "epoch": 1.882379677794051,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00018740291262135919,
      "loss": 0.6033,
      "step": 4660
    },
    {
      "epoch": 1.8827836977930408,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018737864077669902,
      "loss": 0.6701,
      "step": 4661
    },
    {
      "epoch": 1.8831877177920306,
      "grad_norm": 0.484375,
      "learning_rate": 0.00018735436893203882,
      "loss": 0.5988,
      "step": 4662
    },
    {
      "epoch": 1.8835917377910207,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00018733009708737865,
      "loss": 0.6473,
      "step": 4663
    },
    {
      "epoch": 1.8839957577900106,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00018730582524271842,
      "loss": 0.5677,
      "step": 4664
    },
    {
      "epoch": 1.8843997777890005,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00018728155339805822,
      "loss": 0.5491,
      "step": 4665
    },
    {
      "epoch": 1.8848037977879906,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00018725728155339805,
      "loss": 0.6139,
      "step": 4666
    },
    {
      "epoch": 1.8852078177869804,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00018723300970873786,
      "loss": 0.6144,
      "step": 4667
    },
    {
      "epoch": 1.8856118377859703,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018720873786407763,
      "loss": 0.6216,
      "step": 4668
    },
    {
      "epoch": 1.8860158577849604,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00018718446601941746,
      "loss": 0.6412,
      "step": 4669
    },
    {
      "epoch": 1.8864198777839503,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018716019417475726,
      "loss": 0.5961,
      "step": 4670
    },
    {
      "epoch": 1.8868238977829401,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00018713592233009707,
      "loss": 0.5669,
      "step": 4671
    },
    {
      "epoch": 1.8872279177819302,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00018711165048543687,
      "loss": 0.6199,
      "step": 4672
    },
    {
      "epoch": 1.8876319377809203,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00018708737864077667,
      "loss": 0.6312,
      "step": 4673
    },
    {
      "epoch": 1.88803595777991,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0001870631067961165,
      "loss": 0.6905,
      "step": 4674
    },
    {
      "epoch": 1.8884399777789,
      "grad_norm": 0.4921875,
      "learning_rate": 0.0001870388349514563,
      "loss": 0.6435,
      "step": 4675
    },
    {
      "epoch": 1.8888439977778901,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018701456310679608,
      "loss": 0.6576,
      "step": 4676
    },
    {
      "epoch": 1.88924801777688,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0001869902912621359,
      "loss": 0.6391,
      "step": 4677
    },
    {
      "epoch": 1.8896520377758699,
      "grad_norm": 0.546875,
      "learning_rate": 0.0001869660194174757,
      "loss": 0.6528,
      "step": 4678
    },
    {
      "epoch": 1.89005605777486,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0001869417475728155,
      "loss": 0.667,
      "step": 4679
    },
    {
      "epoch": 1.8904600777738498,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00018691747572815534,
      "loss": 0.7255,
      "step": 4680
    },
    {
      "epoch": 1.8908640977728397,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00018689320388349512,
      "loss": 0.6289,
      "step": 4681
    },
    {
      "epoch": 1.8912681177718298,
      "grad_norm": 0.5,
      "learning_rate": 0.00018686893203883495,
      "loss": 0.5854,
      "step": 4682
    },
    {
      "epoch": 1.8916721377708197,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00018684466019417475,
      "loss": 0.6466,
      "step": 4683
    },
    {
      "epoch": 1.8920761577698095,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00018682038834951455,
      "loss": 0.5807,
      "step": 4684
    },
    {
      "epoch": 1.8924801777687996,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018679611650485435,
      "loss": 0.5473,
      "step": 4685
    },
    {
      "epoch": 1.8928841977677895,
      "grad_norm": 0.5,
      "learning_rate": 0.00018677184466019416,
      "loss": 0.6322,
      "step": 4686
    },
    {
      "epoch": 1.8932882177667794,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00018674757281553396,
      "loss": 0.5954,
      "step": 4687
    },
    {
      "epoch": 1.8936922377657694,
      "grad_norm": 0.49609375,
      "learning_rate": 0.0001867233009708738,
      "loss": 0.6541,
      "step": 4688
    },
    {
      "epoch": 1.8940962577647593,
      "grad_norm": 0.4140625,
      "learning_rate": 0.00018669902912621356,
      "loss": 0.5511,
      "step": 4689
    },
    {
      "epoch": 1.8945002777637492,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018667475728155336,
      "loss": 0.632,
      "step": 4690
    },
    {
      "epoch": 1.8949042977627393,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0001866504854368932,
      "loss": 0.583,
      "step": 4691
    },
    {
      "epoch": 1.8953083177617294,
      "grad_norm": 0.58984375,
      "learning_rate": 0.000186626213592233,
      "loss": 0.641,
      "step": 4692
    },
    {
      "epoch": 1.895712337760719,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00018660194174757283,
      "loss": 0.5568,
      "step": 4693
    },
    {
      "epoch": 1.896116357759709,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0001865776699029126,
      "loss": 0.5496,
      "step": 4694
    },
    {
      "epoch": 1.8965203777586992,
      "grad_norm": 0.46875,
      "learning_rate": 0.0001865533980582524,
      "loss": 0.5856,
      "step": 4695
    },
    {
      "epoch": 1.8969243977576888,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00018652912621359223,
      "loss": 0.6079,
      "step": 4696
    },
    {
      "epoch": 1.897328417756679,
      "grad_norm": 0.60546875,
      "learning_rate": 0.000186504854368932,
      "loss": 0.6647,
      "step": 4697
    },
    {
      "epoch": 1.897732437755669,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0001864805825242718,
      "loss": 0.5367,
      "step": 4698
    },
    {
      "epoch": 1.898136457754659,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00018645631067961164,
      "loss": 0.5719,
      "step": 4699
    },
    {
      "epoch": 1.8985404777536488,
      "grad_norm": 0.890625,
      "learning_rate": 0.00018643203883495144,
      "loss": 0.7013,
      "step": 4700
    },
    {
      "epoch": 1.8989444977526388,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00018640776699029122,
      "loss": 0.5949,
      "step": 4701
    },
    {
      "epoch": 1.8993485177516287,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00018638349514563105,
      "loss": 0.5951,
      "step": 4702
    },
    {
      "epoch": 1.8997525377506186,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00018635922330097085,
      "loss": 0.58,
      "step": 4703
    },
    {
      "epoch": 1.9001565577496087,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018633495145631068,
      "loss": 0.6089,
      "step": 4704
    },
    {
      "epoch": 1.9005605777485985,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00018631067961165048,
      "loss": 0.6494,
      "step": 4705
    },
    {
      "epoch": 1.9009645977475884,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00018628640776699026,
      "loss": 0.6633,
      "step": 4706
    },
    {
      "epoch": 1.9013686177465785,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00018626213592233009,
      "loss": 0.6249,
      "step": 4707
    },
    {
      "epoch": 1.9017726377455684,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0001862378640776699,
      "loss": 0.7218,
      "step": 4708
    },
    {
      "epoch": 1.9021766577445582,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0001862135922330097,
      "loss": 0.6133,
      "step": 4709
    },
    {
      "epoch": 1.9025806777435483,
      "grad_norm": 0.453125,
      "learning_rate": 0.0001861893203883495,
      "loss": 0.5615,
      "step": 4710
    },
    {
      "epoch": 1.9029846977425382,
      "grad_norm": 0.5625,
      "learning_rate": 0.0001861650485436893,
      "loss": 0.6608,
      "step": 4711
    },
    {
      "epoch": 1.903388717741528,
      "grad_norm": 0.53125,
      "learning_rate": 0.00018614077669902912,
      "loss": 0.6088,
      "step": 4712
    },
    {
      "epoch": 1.9037927377405182,
      "grad_norm": 0.41796875,
      "learning_rate": 0.00018611650485436893,
      "loss": 0.5175,
      "step": 4713
    },
    {
      "epoch": 1.9041967577395083,
      "grad_norm": 0.53125,
      "learning_rate": 0.0001860922330097087,
      "loss": 0.5677,
      "step": 4714
    },
    {
      "epoch": 1.904600777738498,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018606796116504853,
      "loss": 0.5513,
      "step": 4715
    },
    {
      "epoch": 1.905004797737488,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00018604368932038833,
      "loss": 0.5619,
      "step": 4716
    },
    {
      "epoch": 1.905408817736478,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00018601941747572814,
      "loss": 0.5667,
      "step": 4717
    },
    {
      "epoch": 1.905812837735468,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00018599514563106797,
      "loss": 0.628,
      "step": 4718
    },
    {
      "epoch": 1.9062168577344578,
      "grad_norm": 0.546875,
      "learning_rate": 0.00018597087378640774,
      "loss": 0.6496,
      "step": 4719
    },
    {
      "epoch": 1.906620877733448,
      "grad_norm": 0.4609375,
      "learning_rate": 0.00018594660194174754,
      "loss": 0.5561,
      "step": 4720
    },
    {
      "epoch": 1.9070248977324378,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00018592233009708737,
      "loss": 0.5854,
      "step": 4721
    },
    {
      "epoch": 1.9074289177314276,
      "grad_norm": 0.43359375,
      "learning_rate": 0.00018589805825242718,
      "loss": 0.6027,
      "step": 4722
    },
    {
      "epoch": 1.9078329377304177,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00018587378640776698,
      "loss": 0.5959,
      "step": 4723
    },
    {
      "epoch": 1.9082369577294076,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00018584951456310678,
      "loss": 0.6913,
      "step": 4724
    },
    {
      "epoch": 1.9086409777283975,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00018582524271844658,
      "loss": 0.6813,
      "step": 4725
    },
    {
      "epoch": 1.9090449977273876,
      "grad_norm": 0.423828125,
      "learning_rate": 0.0001858009708737864,
      "loss": 0.5392,
      "step": 4726
    },
    {
      "epoch": 1.9094490177263774,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0001857766990291262,
      "loss": 0.6317,
      "step": 4727
    },
    {
      "epoch": 1.9098530377253673,
      "grad_norm": 0.6015625,
      "learning_rate": 0.000185752427184466,
      "loss": 0.7168,
      "step": 4728
    },
    {
      "epoch": 1.9102570577243574,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00018572815533980582,
      "loss": 0.606,
      "step": 4729
    },
    {
      "epoch": 1.9106610777233473,
      "grad_norm": 0.578125,
      "learning_rate": 0.00018570388349514562,
      "loss": 0.687,
      "step": 4730
    },
    {
      "epoch": 1.9110650977223371,
      "grad_norm": 0.5390625,
      "learning_rate": 0.0001856796116504854,
      "loss": 0.609,
      "step": 4731
    },
    {
      "epoch": 1.9114691177213272,
      "grad_norm": 0.453125,
      "learning_rate": 0.00018565533980582523,
      "loss": 0.5683,
      "step": 4732
    },
    {
      "epoch": 1.9118731377203173,
      "grad_norm": 0.56640625,
      "learning_rate": 0.00018563106796116503,
      "loss": 0.6262,
      "step": 4733
    },
    {
      "epoch": 1.912277157719307,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00018560679611650486,
      "loss": 0.6142,
      "step": 4734
    },
    {
      "epoch": 1.912681177718297,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00018558252427184463,
      "loss": 0.599,
      "step": 4735
    },
    {
      "epoch": 1.9130851977172871,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00018555825242718444,
      "loss": 0.5484,
      "step": 4736
    },
    {
      "epoch": 1.913489217716277,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018553398058252426,
      "loss": 0.5258,
      "step": 4737
    },
    {
      "epoch": 1.9138932377152669,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018550970873786407,
      "loss": 0.5991,
      "step": 4738
    },
    {
      "epoch": 1.914297257714257,
      "grad_norm": 0.640625,
      "learning_rate": 0.00018548543689320384,
      "loss": 0.7087,
      "step": 4739
    },
    {
      "epoch": 1.9147012777132468,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00018546116504854367,
      "loss": 0.584,
      "step": 4740
    },
    {
      "epoch": 1.9151052977122367,
      "grad_norm": 0.546875,
      "learning_rate": 0.00018543689320388347,
      "loss": 0.5925,
      "step": 4741
    },
    {
      "epoch": 1.9155093177112268,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0001854126213592233,
      "loss": 0.5568,
      "step": 4742
    },
    {
      "epoch": 1.9159133377102167,
      "grad_norm": 0.453125,
      "learning_rate": 0.0001853883495145631,
      "loss": 0.6234,
      "step": 4743
    },
    {
      "epoch": 1.9163173577092065,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00018536407766990288,
      "loss": 0.6529,
      "step": 4744
    },
    {
      "epoch": 1.9167213777081966,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0001853398058252427,
      "loss": 0.6019,
      "step": 4745
    },
    {
      "epoch": 1.9171253977071865,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0001853155339805825,
      "loss": 0.574,
      "step": 4746
    },
    {
      "epoch": 1.9175294177061764,
      "grad_norm": 0.46484375,
      "learning_rate": 0.00018529126213592232,
      "loss": 0.6006,
      "step": 4747
    },
    {
      "epoch": 1.9179334377051664,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00018526699029126212,
      "loss": 0.5921,
      "step": 4748
    },
    {
      "epoch": 1.9183374577041563,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00018524271844660192,
      "loss": 0.585,
      "step": 4749
    },
    {
      "epoch": 1.9187414777031462,
      "grad_norm": 0.734375,
      "learning_rate": 0.00018521844660194172,
      "loss": 0.657,
      "step": 4750
    },
    {
      "epoch": 1.9191454977021363,
      "grad_norm": 0.447265625,
      "learning_rate": 0.00018519417475728155,
      "loss": 0.5993,
      "step": 4751
    },
    {
      "epoch": 1.9195495177011264,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00018516990291262133,
      "loss": 0.6926,
      "step": 4752
    },
    {
      "epoch": 1.919953537700116,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00018514563106796116,
      "loss": 0.5514,
      "step": 4753
    },
    {
      "epoch": 1.920357557699106,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00018512135922330096,
      "loss": 0.5898,
      "step": 4754
    },
    {
      "epoch": 1.9207615776980962,
      "grad_norm": 0.5,
      "learning_rate": 0.00018509708737864076,
      "loss": 0.6821,
      "step": 4755
    },
    {
      "epoch": 1.921165597697086,
      "grad_norm": 0.451171875,
      "learning_rate": 0.0001850728155339806,
      "loss": 0.5102,
      "step": 4756
    },
    {
      "epoch": 1.921569617696076,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018504854368932037,
      "loss": 0.6354,
      "step": 4757
    },
    {
      "epoch": 1.921973637695066,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018502427184466017,
      "loss": 0.6232,
      "step": 4758
    },
    {
      "epoch": 1.9223776576940559,
      "grad_norm": 0.46484375,
      "learning_rate": 0.000185,
      "loss": 0.6067,
      "step": 4759
    },
    {
      "epoch": 1.9227816776930458,
      "grad_norm": 0.42578125,
      "learning_rate": 0.0001849757281553398,
      "loss": 0.5676,
      "step": 4760
    },
    {
      "epoch": 1.9231856976920358,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00018495145631067957,
      "loss": 0.6196,
      "step": 4761
    },
    {
      "epoch": 1.9235897176910257,
      "grad_norm": 0.470703125,
      "learning_rate": 0.0001849271844660194,
      "loss": 0.6366,
      "step": 4762
    },
    {
      "epoch": 1.9239937376900156,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0001849029126213592,
      "loss": 0.5936,
      "step": 4763
    },
    {
      "epoch": 1.9243977576890057,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00018487864077669904,
      "loss": 0.6008,
      "step": 4764
    },
    {
      "epoch": 1.9248017776879955,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0001848543689320388,
      "loss": 0.7194,
      "step": 4765
    },
    {
      "epoch": 1.9252057976869854,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00018483009708737861,
      "loss": 0.6609,
      "step": 4766
    },
    {
      "epoch": 1.9256098176859755,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00018480582524271844,
      "loss": 0.5393,
      "step": 4767
    },
    {
      "epoch": 1.9260138376849654,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00018478155339805825,
      "loss": 0.6052,
      "step": 4768
    },
    {
      "epoch": 1.9264178576839552,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00018475728155339802,
      "loss": 0.6683,
      "step": 4769
    },
    {
      "epoch": 1.9268218776829453,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00018473300970873785,
      "loss": 0.5462,
      "step": 4770
    },
    {
      "epoch": 1.9272258976819354,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00018470873786407765,
      "loss": 0.6946,
      "step": 4771
    },
    {
      "epoch": 1.927629917680925,
      "grad_norm": 0.625,
      "learning_rate": 0.00018468446601941745,
      "loss": 0.7003,
      "step": 4772
    },
    {
      "epoch": 1.9280339376799152,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00018466019417475726,
      "loss": 0.7,
      "step": 4773
    },
    {
      "epoch": 1.9284379576789052,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00018463592233009706,
      "loss": 0.551,
      "step": 4774
    },
    {
      "epoch": 1.928841977677895,
      "grad_norm": 0.56640625,
      "learning_rate": 0.0001846116504854369,
      "loss": 0.6524,
      "step": 4775
    },
    {
      "epoch": 1.929245997676885,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0001845873786407767,
      "loss": 0.606,
      "step": 4776
    },
    {
      "epoch": 1.929650017675875,
      "grad_norm": 0.44140625,
      "learning_rate": 0.00018456310679611647,
      "loss": 0.5033,
      "step": 4777
    },
    {
      "epoch": 1.930054037674865,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0001845388349514563,
      "loss": 0.6652,
      "step": 4778
    },
    {
      "epoch": 1.9304580576738548,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0001845145631067961,
      "loss": 0.5714,
      "step": 4779
    },
    {
      "epoch": 1.930862077672845,
      "grad_norm": 0.515625,
      "learning_rate": 0.0001844902912621359,
      "loss": 0.6326,
      "step": 4780
    },
    {
      "epoch": 1.9312660976718348,
      "grad_norm": 0.5,
      "learning_rate": 0.00018446601941747573,
      "loss": 0.5858,
      "step": 4781
    },
    {
      "epoch": 1.9316701176708246,
      "grad_norm": 0.51171875,
      "learning_rate": 0.0001844417475728155,
      "loss": 0.6605,
      "step": 4782
    },
    {
      "epoch": 1.9320741376698147,
      "grad_norm": 0.42578125,
      "learning_rate": 0.00018441747572815533,
      "loss": 0.5641,
      "step": 4783
    },
    {
      "epoch": 1.9324781576688046,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00018439320388349514,
      "loss": 0.571,
      "step": 4784
    },
    {
      "epoch": 1.9328821776677945,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018436893203883494,
      "loss": 0.6421,
      "step": 4785
    },
    {
      "epoch": 1.9332861976667846,
      "grad_norm": 0.4375,
      "learning_rate": 0.00018434466019417474,
      "loss": 0.6435,
      "step": 4786
    },
    {
      "epoch": 1.9336902176657744,
      "grad_norm": 0.53125,
      "learning_rate": 0.00018432038834951454,
      "loss": 0.6207,
      "step": 4787
    },
    {
      "epoch": 1.9340942376647643,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00018429611650485435,
      "loss": 0.5997,
      "step": 4788
    },
    {
      "epoch": 1.9344982576637544,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00018427184466019418,
      "loss": 0.6071,
      "step": 4789
    },
    {
      "epoch": 1.9349022776627443,
      "grad_norm": 0.5625,
      "learning_rate": 0.00018424757281553395,
      "loss": 0.712,
      "step": 4790
    },
    {
      "epoch": 1.9353062976617341,
      "grad_norm": 0.46875,
      "learning_rate": 0.00018422330097087375,
      "loss": 0.5839,
      "step": 4791
    },
    {
      "epoch": 1.9357103176607242,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018419902912621358,
      "loss": 0.5827,
      "step": 4792
    },
    {
      "epoch": 1.9361143376597143,
      "grad_norm": 0.439453125,
      "learning_rate": 0.00018417475728155339,
      "loss": 0.5612,
      "step": 4793
    },
    {
      "epoch": 1.936518357658704,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018415048543689321,
      "loss": 0.6506,
      "step": 4794
    },
    {
      "epoch": 1.936922377657694,
      "grad_norm": 0.51953125,
      "learning_rate": 0.000184126213592233,
      "loss": 0.6051,
      "step": 4795
    },
    {
      "epoch": 1.9373263976566841,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0001841019417475728,
      "loss": 0.6074,
      "step": 4796
    },
    {
      "epoch": 1.937730417655674,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00018407766990291262,
      "loss": 0.5777,
      "step": 4797
    },
    {
      "epoch": 1.9381344376546639,
      "grad_norm": 0.478515625,
      "learning_rate": 0.0001840533980582524,
      "loss": 0.5693,
      "step": 4798
    },
    {
      "epoch": 1.938538457653654,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0001840291262135922,
      "loss": 0.7221,
      "step": 4799
    },
    {
      "epoch": 1.9389424776526438,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00018400485436893203,
      "loss": 0.585,
      "step": 4800
    },
    {
      "epoch": 1.9393464976516337,
      "grad_norm": 0.5,
      "learning_rate": 0.00018398058252427183,
      "loss": 0.5923,
      "step": 4801
    },
    {
      "epoch": 1.9397505176506238,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0001839563106796116,
      "loss": 0.6124,
      "step": 4802
    },
    {
      "epoch": 1.9401545376496137,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018393203883495144,
      "loss": 0.6275,
      "step": 4803
    },
    {
      "epoch": 1.9405585576486035,
      "grad_norm": 0.486328125,
      "learning_rate": 0.00018390776699029124,
      "loss": 0.5879,
      "step": 4804
    },
    {
      "epoch": 1.9409625776475936,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00018388349514563107,
      "loss": 0.7243,
      "step": 4805
    },
    {
      "epoch": 1.9413665976465835,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00018385922330097087,
      "loss": 0.5489,
      "step": 4806
    },
    {
      "epoch": 1.9417706176455733,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00018383495145631065,
      "loss": 0.5793,
      "step": 4807
    },
    {
      "epoch": 1.9421746376445634,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00018381067961165047,
      "loss": 0.6101,
      "step": 4808
    },
    {
      "epoch": 1.9425786576435533,
      "grad_norm": 0.4453125,
      "learning_rate": 0.00018378640776699028,
      "loss": 0.5413,
      "step": 4809
    },
    {
      "epoch": 1.9429826776425432,
      "grad_norm": 0.5625,
      "learning_rate": 0.00018376213592233008,
      "loss": 0.5781,
      "step": 4810
    },
    {
      "epoch": 1.9433866976415333,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00018373786407766988,
      "loss": 0.6248,
      "step": 4811
    },
    {
      "epoch": 1.9437907176405234,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00018371359223300968,
      "loss": 0.577,
      "step": 4812
    },
    {
      "epoch": 1.944194737639513,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0001836893203883495,
      "loss": 0.6692,
      "step": 4813
    },
    {
      "epoch": 1.944598757638503,
      "grad_norm": 0.484375,
      "learning_rate": 0.00018366504854368932,
      "loss": 0.5554,
      "step": 4814
    },
    {
      "epoch": 1.9450027776374932,
      "grad_norm": 0.466796875,
      "learning_rate": 0.0001836407766990291,
      "loss": 0.5911,
      "step": 4815
    },
    {
      "epoch": 1.945406797636483,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00018361650485436892,
      "loss": 0.6242,
      "step": 4816
    },
    {
      "epoch": 1.945810817635473,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00018359223300970872,
      "loss": 0.6061,
      "step": 4817
    },
    {
      "epoch": 1.946214837634463,
      "grad_norm": 0.578125,
      "learning_rate": 0.00018356796116504853,
      "loss": 0.6348,
      "step": 4818
    },
    {
      "epoch": 1.9466188576334529,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00018354368932038835,
      "loss": 0.5995,
      "step": 4819
    },
    {
      "epoch": 1.9470228776324427,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018351941747572813,
      "loss": 0.7065,
      "step": 4820
    },
    {
      "epoch": 1.9474268976314328,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00018349514563106793,
      "loss": 0.5193,
      "step": 4821
    },
    {
      "epoch": 1.9478309176304227,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00018347087378640776,
      "loss": 0.5875,
      "step": 4822
    },
    {
      "epoch": 1.9482349376294126,
      "grad_norm": 0.640625,
      "learning_rate": 0.00018344660194174756,
      "loss": 0.6635,
      "step": 4823
    },
    {
      "epoch": 1.9486389576284027,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00018342233009708737,
      "loss": 0.6571,
      "step": 4824
    },
    {
      "epoch": 1.9490429776273925,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018339805825242717,
      "loss": 0.6176,
      "step": 4825
    },
    {
      "epoch": 1.9494469976263824,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00018337378640776697,
      "loss": 0.5891,
      "step": 4826
    },
    {
      "epoch": 1.9498510176253725,
      "grad_norm": 0.53125,
      "learning_rate": 0.0001833495145631068,
      "loss": 0.6101,
      "step": 4827
    },
    {
      "epoch": 1.9502550376243624,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00018332524271844658,
      "loss": 0.6648,
      "step": 4828
    },
    {
      "epoch": 1.9506590576233522,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00018330097087378638,
      "loss": 0.6068,
      "step": 4829
    },
    {
      "epoch": 1.9510630776223423,
      "grad_norm": 0.44921875,
      "learning_rate": 0.0001832766990291262,
      "loss": 0.5734,
      "step": 4830
    },
    {
      "epoch": 1.9514670976213324,
      "grad_norm": 0.45703125,
      "learning_rate": 0.000183252427184466,
      "loss": 0.589,
      "step": 4831
    },
    {
      "epoch": 1.951871117620322,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018322815533980579,
      "loss": 0.6016,
      "step": 4832
    },
    {
      "epoch": 1.9522751376193122,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00018320388349514561,
      "loss": 0.7619,
      "step": 4833
    },
    {
      "epoch": 1.9526791576183022,
      "grad_norm": 0.5546875,
      "learning_rate": 0.00018317961165048542,
      "loss": 0.6835,
      "step": 4834
    },
    {
      "epoch": 1.953083177617292,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018315533980582525,
      "loss": 0.6371,
      "step": 4835
    },
    {
      "epoch": 1.953487197616282,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00018313106796116502,
      "loss": 0.5932,
      "step": 4836
    },
    {
      "epoch": 1.953891217615272,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00018310679611650482,
      "loss": 0.5911,
      "step": 4837
    },
    {
      "epoch": 1.954295237614262,
      "grad_norm": 0.52734375,
      "learning_rate": 0.00018308252427184465,
      "loss": 0.6284,
      "step": 4838
    },
    {
      "epoch": 1.9546992576132518,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00018305825242718446,
      "loss": 0.6824,
      "step": 4839
    },
    {
      "epoch": 1.955103277612242,
      "grad_norm": 0.546875,
      "learning_rate": 0.00018303398058252423,
      "loss": 0.6757,
      "step": 4840
    },
    {
      "epoch": 1.9555072976112318,
      "grad_norm": 0.546875,
      "learning_rate": 0.00018300970873786406,
      "loss": 0.6386,
      "step": 4841
    },
    {
      "epoch": 1.9559113176102216,
      "grad_norm": 0.5625,
      "learning_rate": 0.00018298543689320386,
      "loss": 0.6165,
      "step": 4842
    },
    {
      "epoch": 1.9563153376092117,
      "grad_norm": 0.4453125,
      "learning_rate": 0.0001829611650485437,
      "loss": 0.5488,
      "step": 4843
    },
    {
      "epoch": 1.9567193576082016,
      "grad_norm": 0.55859375,
      "learning_rate": 0.0001829368932038835,
      "loss": 0.6295,
      "step": 4844
    },
    {
      "epoch": 1.9571233776071915,
      "grad_norm": 0.408203125,
      "learning_rate": 0.00018291262135922327,
      "loss": 0.5815,
      "step": 4845
    },
    {
      "epoch": 1.9575273976061816,
      "grad_norm": 0.5703125,
      "learning_rate": 0.0001828883495145631,
      "loss": 0.6677,
      "step": 4846
    },
    {
      "epoch": 1.9579314176051714,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0001828640776699029,
      "loss": 0.626,
      "step": 4847
    },
    {
      "epoch": 1.9583354376041613,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0001828398058252427,
      "loss": 0.5712,
      "step": 4848
    },
    {
      "epoch": 1.9587394576031514,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0001828155339805825,
      "loss": 0.5848,
      "step": 4849
    },
    {
      "epoch": 1.9591434776021415,
      "grad_norm": 0.5,
      "learning_rate": 0.0001827912621359223,
      "loss": 0.5663,
      "step": 4850
    },
    {
      "epoch": 1.9595474976011311,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0001827669902912621,
      "loss": 0.6471,
      "step": 4851
    },
    {
      "epoch": 1.9599515176001212,
      "grad_norm": 0.5,
      "learning_rate": 0.00018274271844660194,
      "loss": 0.6182,
      "step": 4852
    },
    {
      "epoch": 1.9603555375991113,
      "grad_norm": 0.546875,
      "learning_rate": 0.00018271844660194172,
      "loss": 0.6595,
      "step": 4853
    },
    {
      "epoch": 1.960759557598101,
      "grad_norm": 0.578125,
      "learning_rate": 0.00018269417475728154,
      "loss": 0.5994,
      "step": 4854
    },
    {
      "epoch": 1.961163577597091,
      "grad_norm": 0.4921875,
      "learning_rate": 0.00018266990291262135,
      "loss": 0.5898,
      "step": 4855
    },
    {
      "epoch": 1.9615675975960811,
      "grad_norm": 0.5703125,
      "learning_rate": 0.00018264563106796115,
      "loss": 0.5818,
      "step": 4856
    },
    {
      "epoch": 1.961971617595071,
      "grad_norm": 0.47265625,
      "learning_rate": 0.00018262135922330098,
      "loss": 0.5395,
      "step": 4857
    },
    {
      "epoch": 1.9623756375940609,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00018259708737864075,
      "loss": 0.5988,
      "step": 4858
    },
    {
      "epoch": 1.962779657593051,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00018257281553398056,
      "loss": 0.687,
      "step": 4859
    },
    {
      "epoch": 1.9631836775920408,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00018254854368932039,
      "loss": 0.5858,
      "step": 4860
    },
    {
      "epoch": 1.9635876975910307,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018252427184466016,
      "loss": 0.6466,
      "step": 4861
    },
    {
      "epoch": 1.9639917175900208,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00018249999999999996,
      "loss": 0.6215,
      "step": 4862
    },
    {
      "epoch": 1.9643957375890106,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0001824757281553398,
      "loss": 0.6344,
      "step": 4863
    },
    {
      "epoch": 1.9647997575880005,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0001824514563106796,
      "loss": 0.7049,
      "step": 4864
    },
    {
      "epoch": 1.9652037775869906,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00018242718446601942,
      "loss": 0.6539,
      "step": 4865
    },
    {
      "epoch": 1.9656077975859805,
      "grad_norm": 0.54296875,
      "learning_rate": 0.0001824029126213592,
      "loss": 0.6718,
      "step": 4866
    },
    {
      "epoch": 1.9660118175849703,
      "grad_norm": 0.453125,
      "learning_rate": 0.000182378640776699,
      "loss": 0.6545,
      "step": 4867
    },
    {
      "epoch": 1.9664158375839604,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00018235436893203883,
      "loss": 0.6047,
      "step": 4868
    },
    {
      "epoch": 1.9668198575829503,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00018233009708737863,
      "loss": 0.6331,
      "step": 4869
    },
    {
      "epoch": 1.9672238775819402,
      "grad_norm": 0.458984375,
      "learning_rate": 0.0001823058252427184,
      "loss": 0.6036,
      "step": 4870
    },
    {
      "epoch": 1.9676278975809303,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00018228155339805824,
      "loss": 0.6573,
      "step": 4871
    },
    {
      "epoch": 1.9680319175799204,
      "grad_norm": 0.5625,
      "learning_rate": 0.00018225728155339804,
      "loss": 0.6222,
      "step": 4872
    },
    {
      "epoch": 1.96843593757891,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00018223300970873784,
      "loss": 0.5818,
      "step": 4873
    },
    {
      "epoch": 1.9688399575779,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00018220873786407765,
      "loss": 0.5809,
      "step": 4874
    },
    {
      "epoch": 1.9692439775768902,
      "grad_norm": 0.478515625,
      "learning_rate": 0.00018218446601941745,
      "loss": 0.585,
      "step": 4875
    },
    {
      "epoch": 1.96964799757588,
      "grad_norm": 0.50390625,
      "learning_rate": 0.00018216019417475728,
      "loss": 0.594,
      "step": 4876
    },
    {
      "epoch": 1.97005201757487,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018213592233009708,
      "loss": 0.4994,
      "step": 4877
    },
    {
      "epoch": 1.97045603757386,
      "grad_norm": 0.609375,
      "learning_rate": 0.00018211165048543686,
      "loss": 0.5394,
      "step": 4878
    },
    {
      "epoch": 1.9708600575728499,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00018208737864077668,
      "loss": 0.6305,
      "step": 4879
    },
    {
      "epoch": 1.9712640775718397,
      "grad_norm": 0.57421875,
      "learning_rate": 0.0001820631067961165,
      "loss": 0.6272,
      "step": 4880
    },
    {
      "epoch": 1.9716680975708298,
      "grad_norm": 0.44921875,
      "learning_rate": 0.0001820388349514563,
      "loss": 0.5566,
      "step": 4881
    },
    {
      "epoch": 1.9720721175698197,
      "grad_norm": 0.53125,
      "learning_rate": 0.00018201456310679612,
      "loss": 0.6288,
      "step": 4882
    },
    {
      "epoch": 1.9724761375688096,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0001819902912621359,
      "loss": 0.5678,
      "step": 4883
    },
    {
      "epoch": 1.9728801575677997,
      "grad_norm": 0.466796875,
      "learning_rate": 0.00018196601941747572,
      "loss": 0.5828,
      "step": 4884
    },
    {
      "epoch": 1.9732841775667895,
      "grad_norm": 0.484375,
      "learning_rate": 0.00018194174757281553,
      "loss": 0.5643,
      "step": 4885
    },
    {
      "epoch": 1.9736881975657794,
      "grad_norm": 0.6875,
      "learning_rate": 0.00018191747572815533,
      "loss": 0.6031,
      "step": 4886
    },
    {
      "epoch": 1.9740922175647695,
      "grad_norm": 0.609375,
      "learning_rate": 0.00018189320388349513,
      "loss": 0.6955,
      "step": 4887
    },
    {
      "epoch": 1.9744962375637594,
      "grad_norm": 0.5625,
      "learning_rate": 0.00018186893203883493,
      "loss": 0.6263,
      "step": 4888
    },
    {
      "epoch": 1.9749002575627492,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00018184466019417474,
      "loss": 0.6036,
      "step": 4889
    },
    {
      "epoch": 1.9753042775617393,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00018182038834951456,
      "loss": 0.6111,
      "step": 4890
    },
    {
      "epoch": 1.9757082975607294,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00018179611650485434,
      "loss": 0.6762,
      "step": 4891
    },
    {
      "epoch": 1.976112317559719,
      "grad_norm": 0.470703125,
      "learning_rate": 0.00018177184466019414,
      "loss": 0.6036,
      "step": 4892
    },
    {
      "epoch": 1.9765163375587091,
      "grad_norm": 0.453125,
      "learning_rate": 0.00018174757281553397,
      "loss": 0.5552,
      "step": 4893
    },
    {
      "epoch": 1.9769203575576992,
      "grad_norm": 0.455078125,
      "learning_rate": 0.00018172330097087377,
      "loss": 0.5612,
      "step": 4894
    },
    {
      "epoch": 1.977324377556689,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0001816990291262136,
      "loss": 0.5987,
      "step": 4895
    },
    {
      "epoch": 1.977728397555679,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00018167475728155338,
      "loss": 0.5713,
      "step": 4896
    },
    {
      "epoch": 1.978132417554669,
      "grad_norm": 0.51953125,
      "learning_rate": 0.00018165048543689318,
      "loss": 0.6092,
      "step": 4897
    },
    {
      "epoch": 1.978536437553659,
      "grad_norm": 0.52734375,
      "learning_rate": 0.000181626213592233,
      "loss": 0.6406,
      "step": 4898
    },
    {
      "epoch": 1.9789404575526488,
      "grad_norm": 0.609375,
      "learning_rate": 0.00018160194174757279,
      "loss": 0.6589,
      "step": 4899
    },
    {
      "epoch": 1.979344477551639,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0001815776699029126,
      "loss": 0.5505,
      "step": 4900
    },
    {
      "epoch": 1.9797484975506288,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00018155339805825242,
      "loss": 0.5697,
      "step": 4901
    },
    {
      "epoch": 1.9801525175496186,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00018152912621359222,
      "loss": 0.5362,
      "step": 4902
    },
    {
      "epoch": 1.9805565375486087,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000181504854368932,
      "loss": 0.7898,
      "step": 4903
    },
    {
      "epoch": 1.9809605575475986,
      "grad_norm": 0.462890625,
      "learning_rate": 0.00018148058252427182,
      "loss": 0.5807,
      "step": 4904
    },
    {
      "epoch": 1.9813645775465885,
      "grad_norm": 0.44921875,
      "learning_rate": 0.00018145631067961163,
      "loss": 0.5288,
      "step": 4905
    },
    {
      "epoch": 1.9817685975455785,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00018143203883495146,
      "loss": 0.7049,
      "step": 4906
    },
    {
      "epoch": 1.9821726175445684,
      "grad_norm": 0.53515625,
      "learning_rate": 0.00018140776699029126,
      "loss": 0.608,
      "step": 4907
    },
    {
      "epoch": 1.9825766375435583,
      "grad_norm": 0.474609375,
      "learning_rate": 0.00018138349514563103,
      "loss": 0.5033,
      "step": 4908
    },
    {
      "epoch": 1.9829806575425484,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00018135922330097086,
      "loss": 0.6532,
      "step": 4909
    },
    {
      "epoch": 1.9833846775415385,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00018133495145631067,
      "loss": 0.6683,
      "step": 4910
    },
    {
      "epoch": 1.9837886975405281,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00018131067961165047,
      "loss": 0.5844,
      "step": 4911
    },
    {
      "epoch": 1.9841927175395182,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00018128640776699027,
      "loss": 0.655,
      "step": 4912
    },
    {
      "epoch": 1.9845967375385083,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018126213592233007,
      "loss": 0.6468,
      "step": 4913
    },
    {
      "epoch": 1.9850007575374982,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0001812378640776699,
      "loss": 0.6575,
      "step": 4914
    },
    {
      "epoch": 1.985404777536488,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0001812135922330097,
      "loss": 0.6421,
      "step": 4915
    },
    {
      "epoch": 1.9858087975354781,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00018118932038834948,
      "loss": 0.5766,
      "step": 4916
    },
    {
      "epoch": 1.986212817534468,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0001811650485436893,
      "loss": 0.6008,
      "step": 4917
    },
    {
      "epoch": 1.9866168375334579,
      "grad_norm": 0.65234375,
      "learning_rate": 0.0001811407766990291,
      "loss": 0.6701,
      "step": 4918
    },
    {
      "epoch": 1.987020857532448,
      "grad_norm": 0.447265625,
      "learning_rate": 0.00018111650485436891,
      "loss": 0.5376,
      "step": 4919
    },
    {
      "epoch": 1.9874248775314378,
      "grad_norm": 0.6015625,
      "learning_rate": 0.00018109223300970874,
      "loss": 0.5981,
      "step": 4920
    },
    {
      "epoch": 1.9878288975304277,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00018106796116504852,
      "loss": 0.5876,
      "step": 4921
    },
    {
      "epoch": 1.9882329175294178,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018104368932038832,
      "loss": 0.6429,
      "step": 4922
    },
    {
      "epoch": 1.9886369375284076,
      "grad_norm": 0.5078125,
      "learning_rate": 0.00018101941747572815,
      "loss": 0.637,
      "step": 4923
    },
    {
      "epoch": 1.9890409575273975,
      "grad_norm": 0.49609375,
      "learning_rate": 0.00018099514563106795,
      "loss": 0.5877,
      "step": 4924
    },
    {
      "epoch": 1.9894449775263876,
      "grad_norm": 0.51171875,
      "learning_rate": 0.00018097087378640776,
      "loss": 0.6946,
      "step": 4925
    },
    {
      "epoch": 1.9898489975253775,
      "grad_norm": 0.490234375,
      "learning_rate": 0.00018094660194174756,
      "loss": 0.6038,
      "step": 4926
    },
    {
      "epoch": 1.9902530175243673,
      "grad_norm": 0.48828125,
      "learning_rate": 0.00018092233009708736,
      "loss": 0.6368,
      "step": 4927
    },
    {
      "epoch": 1.9906570375233574,
      "grad_norm": 0.515625,
      "learning_rate": 0.0001808980582524272,
      "loss": 0.6769,
      "step": 4928
    },
    {
      "epoch": 1.9910610575223475,
      "grad_norm": 0.447265625,
      "learning_rate": 0.00018087378640776696,
      "loss": 0.5713,
      "step": 4929
    },
    {
      "epoch": 1.9914650775213372,
      "grad_norm": 0.55859375,
      "learning_rate": 0.00018084951456310677,
      "loss": 0.6566,
      "step": 4930
    },
    {
      "epoch": 1.9918690975203273,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0001808252427184466,
      "loss": 0.6129,
      "step": 4931
    },
    {
      "epoch": 1.9922731175193173,
      "grad_norm": 0.578125,
      "learning_rate": 0.0001808009708737864,
      "loss": 0.6452,
      "step": 4932
    },
    {
      "epoch": 1.992677137518307,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00018077669902912617,
      "loss": 0.6099,
      "step": 4933
    },
    {
      "epoch": 1.993081157517297,
      "grad_norm": 0.546875,
      "learning_rate": 0.000180752427184466,
      "loss": 0.6303,
      "step": 4934
    },
    {
      "epoch": 1.9934851775162872,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0001807281553398058,
      "loss": 0.6086,
      "step": 4935
    },
    {
      "epoch": 1.993889197515277,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00018070388349514564,
      "loss": 0.6491,
      "step": 4936
    },
    {
      "epoch": 1.994293217514267,
      "grad_norm": 0.486328125,
      "learning_rate": 0.0001806796116504854,
      "loss": 0.623,
      "step": 4937
    },
    {
      "epoch": 1.994697237513257,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0001806553398058252,
      "loss": 0.6871,
      "step": 4938
    },
    {
      "epoch": 1.9951012575122469,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00018063106796116504,
      "loss": 0.5789,
      "step": 4939
    },
    {
      "epoch": 1.9955052775112367,
      "grad_norm": 0.53125,
      "learning_rate": 0.00018060679611650484,
      "loss": 0.6759,
      "step": 4940
    },
    {
      "epoch": 1.9959092975102268,
      "grad_norm": 0.46875,
      "learning_rate": 0.00018058252427184462,
      "loss": 0.6401,
      "step": 4941
    },
    {
      "epoch": 1.9963133175092167,
      "grad_norm": 0.5625,
      "learning_rate": 0.00018055825242718445,
      "loss": 0.6444,
      "step": 4942
    },
    {
      "epoch": 1.9967173375082066,
      "grad_norm": 0.498046875,
      "learning_rate": 0.00018053398058252425,
      "loss": 0.6397,
      "step": 4943
    },
    {
      "epoch": 1.9971213575071967,
      "grad_norm": 0.55078125,
      "learning_rate": 0.00018050970873786408,
      "loss": 0.571,
      "step": 4944
    },
    {
      "epoch": 1.9975253775061865,
      "grad_norm": 0.515625,
      "learning_rate": 0.00018048543689320388,
      "loss": 0.6403,
      "step": 4945
    },
    {
      "epoch": 1.9979293975051764,
      "grad_norm": 0.5,
      "learning_rate": 0.00018046116504854366,
      "loss": 0.6238,
      "step": 4946
    },
    {
      "epoch": 1.9983334175041665,
      "grad_norm": 0.5078125,
      "learning_rate": 0.0001804368932038835,
      "loss": 0.6702,
      "step": 4947
    },
    {
      "epoch": 1.9987374375031564,
      "grad_norm": 0.48828125,
      "learning_rate": 0.0001804126213592233,
      "loss": 0.6376,
      "step": 4948
    },
    {
      "epoch": 1.9991414575021462,
      "grad_norm": 0.5234375,
      "learning_rate": 0.0001803883495145631,
      "loss": 0.6554,
      "step": 4949
    },
    {
      "epoch": 1.9995454775011363,
      "grad_norm": 0.490234375,
      "learning_rate": 0.0001803640776699029,
      "loss": 0.5481,
      "step": 4950
    },
    {
      "epoch": 1.9999494975001264,
      "grad_norm": 0.482421875,
      "learning_rate": 0.0001803398058252427,
      "loss": 0.5856,
      "step": 4951
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.3125,
      "learning_rate": 0.0001803155339805825,
      "loss": 0.4942,
      "step": 4952
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.4348788261413574,
      "eval_runtime": 213.7191,
      "eval_samples_per_second": 28.739,
      "eval_steps_per_second": 28.739,
      "step": 4952
    },
    {
      "epoch": 2.0,
      "eval_title2sid_loss": 0.7016080617904663,
      "eval_title2sid_runtime": 126.8384,
      "eval_title2sid_samples_per_second": 30.007,
      "eval_title2sid_steps_per_second": 30.007,
      "step": 4952
    },
    {
      "epoch": 2.0,
      "eval_title2sid_loss": 0.7016080617904663,
      "eval_title2sid_runtime": 126.8384,
      "eval_title2sid_samples_per_second": 30.007,
      "eval_title2sid_steps_per_second": 30.007,
      "step": 4952
    },
    {
      "epoch": 2.0,
      "eval_sid2title_loss": 0.12928932905197144,
      "eval_sid2title_runtime": 127.9496,
      "eval_sid2title_samples_per_second": 29.91,
      "eval_sid2title_steps_per_second": 29.91,
      "step": 4952
    },
    {
      "epoch": 2.0,
      "eval_sid2title_loss": 0.12928932905197144,
      "eval_sid2title_runtime": 127.9496,
      "eval_sid2title_samples_per_second": 29.91,
      "eval_sid2title_steps_per_second": 29.91,
      "step": 4952
    }
  ],
  "logging_steps": 1,
  "max_steps": 12380,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.383451815677231e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}