{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998557067962099,
  "eval_steps": 500,
  "global_step": 5197,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.7290843725204468,
      "learning_rate": 1.282051282051282e-06,
      "loss": 2.0923,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.537783145904541,
      "learning_rate": 2.564102564102564e-06,
      "loss": 2.0294,
      "step": 2
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.5585165619850159,
      "learning_rate": 3.846153846153847e-06,
      "loss": 1.9307,
      "step": 3
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.6017361879348755,
      "learning_rate": 5.128205128205128e-06,
      "loss": 1.9635,
      "step": 4
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.5905205607414246,
      "learning_rate": 6.41025641025641e-06,
      "loss": 1.9175,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.6761202812194824,
      "learning_rate": 7.692307692307694e-06,
      "loss": 2.1621,
      "step": 6
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.6903293132781982,
      "learning_rate": 8.974358974358976e-06,
      "loss": 2.1908,
      "step": 7
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.85490083694458,
      "learning_rate": 1.0256410256410256e-05,
      "loss": 2.1386,
      "step": 8
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.0674844980239868,
      "learning_rate": 1.153846153846154e-05,
      "loss": 2.0465,
      "step": 9
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.0674844980239868,
      "learning_rate": 1.153846153846154e-05,
      "loss": 2.2069,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.7962436676025391,
      "learning_rate": 1.282051282051282e-05,
      "loss": 2.0625,
      "step": 11
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.6697447896003723,
      "learning_rate": 1.4102564102564104e-05,
      "loss": 2.1082,
      "step": 12
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.4376391172409058,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 2.1937,
      "step": 13
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.9062362313270569,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.9937,
      "step": 14
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.9200788140296936,
      "learning_rate": 1.794871794871795e-05,
      "loss": 2.2481,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.8441256880760193,
      "learning_rate": 1.923076923076923e-05,
      "loss": 2.3135,
      "step": 16
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.251808762550354,
      "learning_rate": 2.0512820512820512e-05,
      "loss": 2.2414,
      "step": 17
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.7678605914115906,
      "learning_rate": 2.1794871794871795e-05,
      "loss": 2.3481,
      "step": 18
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.095533847808838,
      "learning_rate": 2.307692307692308e-05,
      "loss": 2.2993,
      "step": 19
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.8871117234230042,
      "learning_rate": 2.435897435897436e-05,
      "loss": 2.0338,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.6377506256103516,
      "learning_rate": 2.564102564102564e-05,
      "loss": 2.0361,
      "step": 21
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.631040632724762,
      "learning_rate": 2.6923076923076923e-05,
      "loss": 2.0365,
      "step": 22
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.7686471939086914,
      "learning_rate": 2.8205128205128207e-05,
      "loss": 2.1325,
      "step": 23
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.5957205891609192,
      "learning_rate": 2.948717948717949e-05,
      "loss": 1.961,
      "step": 24
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.6763018369674683,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 1.9462,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6536284685134888,
      "learning_rate": 3.205128205128206e-05,
      "loss": 2.0485,
      "step": 26
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.4939531981945038,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.7862,
      "step": 27
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.793214738368988,
      "learning_rate": 3.461538461538462e-05,
      "loss": 2.1794,
      "step": 28
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5832692980766296,
      "learning_rate": 3.58974358974359e-05,
      "loss": 1.9488,
      "step": 29
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8362488150596619,
      "learning_rate": 3.717948717948718e-05,
      "loss": 1.9762,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5798012018203735,
      "learning_rate": 3.846153846153846e-05,
      "loss": 1.879,
      "step": 31
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.063071370124817,
      "learning_rate": 3.974358974358974e-05,
      "loss": 2.0236,
      "step": 32
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5717532634735107,
      "learning_rate": 4.1025641025641023e-05,
      "loss": 1.6645,
      "step": 33
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8875914812088013,
      "learning_rate": 4.230769230769231e-05,
      "loss": 1.9861,
      "step": 34
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6677761673927307,
      "learning_rate": 4.358974358974359e-05,
      "loss": 1.8448,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.661694347858429,
      "learning_rate": 4.4871794871794874e-05,
      "loss": 1.8701,
      "step": 36
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7201910018920898,
      "learning_rate": 4.615384615384616e-05,
      "loss": 1.7591,
      "step": 37
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.0301069021224976,
      "learning_rate": 4.7435897435897435e-05,
      "loss": 1.8264,
      "step": 38
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6007413268089294,
      "learning_rate": 4.871794871794872e-05,
      "loss": 1.8582,
      "step": 39
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1926769018173218,
      "learning_rate": 5e-05,
      "loss": 1.7362,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.4716361463069916,
      "learning_rate": 5.128205128205128e-05,
      "loss": 1.6427,
      "step": 41
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.49851030111312866,
      "learning_rate": 5.256410256410257e-05,
      "loss": 1.5575,
      "step": 42
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7268093228340149,
      "learning_rate": 5.384615384615385e-05,
      "loss": 1.7747,
      "step": 43
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.48582518100738525,
      "learning_rate": 5.512820512820514e-05,
      "loss": 1.6444,
      "step": 44
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6579967141151428,
      "learning_rate": 5.6410256410256414e-05,
      "loss": 1.6489,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5355501770973206,
      "learning_rate": 5.769230769230769e-05,
      "loss": 1.6549,
      "step": 46
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.47149765491485596,
      "learning_rate": 5.897435897435898e-05,
      "loss": 1.5256,
      "step": 47
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.46308189630508423,
      "learning_rate": 6.025641025641026e-05,
      "loss": 1.5956,
      "step": 48
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.4326660931110382,
      "learning_rate": 6.153846153846155e-05,
      "loss": 1.5382,
      "step": 49
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.35284343361854553,
      "learning_rate": 6.282051282051282e-05,
      "loss": 1.4669,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.39759883284568787,
      "learning_rate": 6.410256410256412e-05,
      "loss": 1.5222,
      "step": 51
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.3526046872138977,
      "learning_rate": 6.538461538461539e-05,
      "loss": 1.5395,
      "step": 52
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.4054192006587982,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.4396,
      "step": 53
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.3377823531627655,
      "learning_rate": 6.794871794871795e-05,
      "loss": 1.4685,
      "step": 54
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.35546207427978516,
      "learning_rate": 6.923076923076924e-05,
      "loss": 1.4582,
      "step": 55
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.2755766212940216,
      "learning_rate": 7.051282051282052e-05,
      "loss": 1.4978,
      "step": 56
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.22551560401916504,
      "learning_rate": 7.17948717948718e-05,
      "loss": 1.4764,
      "step": 57
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6790602803230286,
      "learning_rate": 7.307692307692307e-05,
      "loss": 1.5035,
      "step": 58
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.225571870803833,
      "learning_rate": 7.435897435897436e-05,
      "loss": 1.5409,
      "step": 59
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.2171548753976822,
      "learning_rate": 7.564102564102564e-05,
      "loss": 1.4694,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.203948512673378,
      "learning_rate": 7.692307692307693e-05,
      "loss": 1.4873,
      "step": 61
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.1814311295747757,
      "learning_rate": 7.820512820512821e-05,
      "loss": 1.4081,
      "step": 62
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.17682871222496033,
      "learning_rate": 7.948717948717948e-05,
      "loss": 1.452,
      "step": 63
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.2913624942302704,
      "learning_rate": 8.076923076923078e-05,
      "loss": 1.3826,
      "step": 64
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.1665887087583542,
      "learning_rate": 8.205128205128205e-05,
      "loss": 1.3782,
      "step": 65
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.15831570327281952,
      "learning_rate": 8.333333333333334e-05,
      "loss": 1.4112,
      "step": 66
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.34056439995765686,
      "learning_rate": 8.461538461538461e-05,
      "loss": 1.45,
      "step": 67
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.2191518396139145,
      "learning_rate": 8.58974358974359e-05,
      "loss": 1.472,
      "step": 68
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.17682230472564697,
      "learning_rate": 8.717948717948718e-05,
      "loss": 1.3977,
      "step": 69
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.21128369867801666,
      "learning_rate": 8.846153846153847e-05,
      "loss": 1.4395,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.17617079615592957,
      "learning_rate": 8.974358974358975e-05,
      "loss": 1.3557,
      "step": 71
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.4705723226070404,
      "learning_rate": 9.102564102564103e-05,
      "loss": 1.484,
      "step": 72
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.14154790341854095,
      "learning_rate": 9.230769230769232e-05,
      "loss": 1.3382,
      "step": 73
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.1877722591161728,
      "learning_rate": 9.35897435897436e-05,
      "loss": 1.5164,
      "step": 74
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.16671311855316162,
      "learning_rate": 9.487179487179487e-05,
      "loss": 1.4379,
      "step": 75
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.1357119083404541,
      "learning_rate": 9.615384615384617e-05,
      "loss": 1.2796,
      "step": 76
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.18603332340717316,
      "learning_rate": 9.743589743589744e-05,
      "loss": 1.4431,
      "step": 77
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.165994331240654,
      "learning_rate": 9.871794871794872e-05,
      "loss": 1.5066,
      "step": 78
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.21108615398406982,
      "learning_rate": 0.0001,
      "loss": 1.4397,
      "step": 79
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.19724680483341217,
      "learning_rate": 0.00010128205128205129,
      "loss": 1.3993,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.19967485964298248,
      "learning_rate": 0.00010256410256410256,
      "loss": 1.4048,
      "step": 81
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.17318454384803772,
      "learning_rate": 0.00010384615384615386,
      "loss": 1.3771,
      "step": 82
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1904263198375702,
      "learning_rate": 0.00010512820512820514,
      "loss": 1.4274,
      "step": 83
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1701108068227768,
      "learning_rate": 0.00010641025641025641,
      "loss": 1.3716,
      "step": 84
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1324114352464676,
      "learning_rate": 0.0001076923076923077,
      "loss": 1.2862,
      "step": 85
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.14505237340927124,
      "learning_rate": 0.00010897435897435896,
      "loss": 1.3041,
      "step": 86
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1671178638935089,
      "learning_rate": 0.00011025641025641027,
      "loss": 1.3956,
      "step": 87
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.16628095507621765,
      "learning_rate": 0.00011153846153846154,
      "loss": 1.4218,
      "step": 88
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.14487555623054504,
      "learning_rate": 0.00011282051282051283,
      "loss": 1.2632,
      "step": 89
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.12290135025978088,
      "learning_rate": 0.0001141025641025641,
      "loss": 1.3433,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1500062197446823,
      "learning_rate": 0.00011538461538461538,
      "loss": 1.3856,
      "step": 91
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.19821228086948395,
      "learning_rate": 0.00011666666666666668,
      "loss": 1.3563,
      "step": 92
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.172821044921875,
      "learning_rate": 0.00011794871794871796,
      "loss": 1.3169,
      "step": 93
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1996234953403473,
      "learning_rate": 0.00011923076923076923,
      "loss": 1.3074,
      "step": 94
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.14572983980178833,
      "learning_rate": 0.00012051282051282052,
      "loss": 1.3347,
      "step": 95
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.17105461657047272,
      "learning_rate": 0.00012179487179487179,
      "loss": 1.3681,
      "step": 96
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.12286887317895889,
      "learning_rate": 0.0001230769230769231,
      "loss": 1.2897,
      "step": 97
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.14899210631847382,
      "learning_rate": 0.00012435897435897437,
      "loss": 1.3171,
      "step": 98
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.20335336029529572,
      "learning_rate": 0.00012564102564102564,
      "loss": 1.3392,
      "step": 99
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.12343753129243851,
      "learning_rate": 0.00012692307692307693,
      "loss": 1.2847,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.15760299563407898,
      "learning_rate": 0.00012820512820512823,
      "loss": 1.3112,
      "step": 101
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.19490450620651245,
      "learning_rate": 0.0001294871794871795,
      "loss": 1.2784,
      "step": 102
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.14551898837089539,
      "learning_rate": 0.00013076923076923077,
      "loss": 1.307,
      "step": 103
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.18424178659915924,
      "learning_rate": 0.00013205128205128204,
      "loss": 1.3713,
      "step": 104
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.11920896172523499,
      "learning_rate": 0.00013333333333333334,
      "loss": 1.2981,
      "step": 105
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1593123823404312,
      "learning_rate": 0.00013461538461538464,
      "loss": 1.3338,
      "step": 106
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1387111246585846,
      "learning_rate": 0.0001358974358974359,
      "loss": 1.2958,
      "step": 107
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.13875919580459595,
      "learning_rate": 0.00013717948717948718,
      "loss": 1.3649,
      "step": 108
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1628897786140442,
      "learning_rate": 0.00013846153846153847,
      "loss": 1.3002,
      "step": 109
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.17144213616847992,
      "learning_rate": 0.00013974358974358974,
      "loss": 1.3194,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.18644458055496216,
      "learning_rate": 0.00014102564102564104,
      "loss": 1.2768,
      "step": 111
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1627991497516632,
      "learning_rate": 0.0001423076923076923,
      "loss": 1.2826,
      "step": 112
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.16085515916347504,
      "learning_rate": 0.0001435897435897436,
      "loss": 1.3054,
      "step": 113
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1542607843875885,
      "learning_rate": 0.00014487179487179488,
      "loss": 1.288,
      "step": 114
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.12633778154850006,
      "learning_rate": 0.00014615384615384615,
      "loss": 1.3121,
      "step": 115
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.16368913650512695,
      "learning_rate": 0.00014743589743589745,
      "loss": 1.2561,
      "step": 116
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.15062902867794037,
      "learning_rate": 0.00014871794871794872,
      "loss": 1.2308,
      "step": 117
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.10931210219860077,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.216,
      "step": 118
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.13495436310768127,
      "learning_rate": 0.00015128205128205128,
      "loss": 1.3256,
      "step": 119
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.17220571637153625,
      "learning_rate": 0.00015256410256410255,
      "loss": 1.3925,
      "step": 120
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.14204230904579163,
      "learning_rate": 0.00015384615384615385,
      "loss": 1.3119,
      "step": 121
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.11951388418674469,
      "learning_rate": 0.00015512820512820515,
      "loss": 1.2501,
      "step": 122
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.10945325344800949,
      "learning_rate": 0.00015641025641025642,
      "loss": 1.2892,
      "step": 123
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.2147054523229599,
      "learning_rate": 0.0001576923076923077,
      "loss": 1.2614,
      "step": 124
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1485561579465866,
      "learning_rate": 0.00015897435897435896,
      "loss": 1.3165,
      "step": 125
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.16703079640865326,
      "learning_rate": 0.00016025641025641028,
      "loss": 1.2176,
      "step": 126
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.1256359964609146,
      "learning_rate": 0.00016153846153846155,
      "loss": 1.3554,
      "step": 127
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.13788390159606934,
      "learning_rate": 0.00016282051282051282,
      "loss": 1.3521,
      "step": 128
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.13277536630630493,
      "learning_rate": 0.0001641025641025641,
      "loss": 1.3563,
      "step": 129
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.20635223388671875,
      "learning_rate": 0.0001653846153846154,
      "loss": 1.3057,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.18450675904750824,
      "learning_rate": 0.0001666666666666667,
      "loss": 1.302,
      "step": 131
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.19233326613903046,
      "learning_rate": 0.00016794871794871796,
      "loss": 1.2414,
      "step": 132
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.14502304792404175,
      "learning_rate": 0.00016923076923076923,
      "loss": 1.3211,
      "step": 133
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.14821577072143555,
      "learning_rate": 0.00017051282051282053,
      "loss": 1.2407,
      "step": 134
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.12206808477640152,
      "learning_rate": 0.0001717948717948718,
      "loss": 1.2694,
      "step": 135
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.11955923587083817,
      "learning_rate": 0.0001730769230769231,
      "loss": 1.1455,
      "step": 136
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1422533243894577,
      "learning_rate": 0.00017435897435897436,
      "loss": 1.2747,
      "step": 137
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.14319650828838348,
      "learning_rate": 0.00017564102564102566,
      "loss": 1.3954,
      "step": 138
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.17978490889072418,
      "learning_rate": 0.00017692307692307693,
      "loss": 1.2991,
      "step": 139
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.15294784307479858,
      "learning_rate": 0.00017820512820512823,
      "loss": 1.2854,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.15269735455513,
      "learning_rate": 0.0001794871794871795,
      "loss": 1.2761,
      "step": 141
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1367342472076416,
      "learning_rate": 0.00018076923076923077,
      "loss": 1.3112,
      "step": 142
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.15280239284038544,
      "learning_rate": 0.00018205128205128207,
      "loss": 1.2952,
      "step": 143
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.14964458346366882,
      "learning_rate": 0.00018333333333333334,
      "loss": 1.3002,
      "step": 144
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.16681908071041107,
      "learning_rate": 0.00018461538461538463,
      "loss": 1.2321,
      "step": 145
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1635741889476776,
      "learning_rate": 0.0001858974358974359,
      "loss": 1.1561,
      "step": 146
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1574012190103531,
      "learning_rate": 0.0001871794871794872,
      "loss": 1.3205,
      "step": 147
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1866392344236374,
      "learning_rate": 0.00018846153846153847,
      "loss": 1.3252,
      "step": 148
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.14143535494804382,
      "learning_rate": 0.00018974358974358974,
      "loss": 1.214,
      "step": 149
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1654847413301468,
      "learning_rate": 0.00019102564102564104,
      "loss": 1.3009,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.12701471149921417,
      "learning_rate": 0.00019230769230769233,
      "loss": 1.2212,
      "step": 151
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.13368256390094757,
      "learning_rate": 0.0001935897435897436,
      "loss": 1.1814,
      "step": 152
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.24185074865818024,
      "learning_rate": 0.00019487179487179487,
      "loss": 1.3277,
      "step": 153
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1536712646484375,
      "learning_rate": 0.00019615384615384615,
      "loss": 1.2118,
      "step": 154
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.14595501124858856,
      "learning_rate": 0.00019743589743589744,
      "loss": 1.2162,
      "step": 155
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.12828391790390015,
      "learning_rate": 0.00019871794871794874,
      "loss": 1.261,
      "step": 156
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.15513499081134796,
      "learning_rate": 0.0002,
      "loss": 1.2966,
      "step": 157
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1914309412240982,
      "learning_rate": 0.00019999998058057615,
      "loss": 1.4174,
      "step": 158
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.2157302051782608,
      "learning_rate": 0.00019999992232231216,
      "loss": 1.3287,
      "step": 159
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.16976693272590637,
      "learning_rate": 0.0001999998252252306,
      "loss": 1.2239,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.28809431195259094,
      "learning_rate": 0.00019999968928936926,
      "loss": 1.2278,
      "step": 161
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.20075291395187378,
      "learning_rate": 0.00019999951451478087,
      "loss": 1.271,
      "step": 162
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.17645922303199768,
      "learning_rate": 0.00019999930090153334,
      "loss": 1.2123,
      "step": 163
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.14547890424728394,
      "learning_rate": 0.00019999904844970962,
      "loss": 1.2506,
      "step": 164
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.14498932659626007,
      "learning_rate": 0.00019999875715940782,
      "loss": 1.1912,
      "step": 165
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1376400738954544,
      "learning_rate": 0.000199998427030741,
      "loss": 1.2401,
      "step": 166
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1343030035495758,
      "learning_rate": 0.00019999805806383738,
      "loss": 1.2702,
      "step": 167
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.12494877725839615,
      "learning_rate": 0.0001999976502588403,
      "loss": 1.2829,
      "step": 168
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1490629017353058,
      "learning_rate": 0.0001999972036159081,
      "loss": 1.2745,
      "step": 169
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.13565689325332642,
      "learning_rate": 0.00019999671813521435,
      "loss": 1.2227,
      "step": 170
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.15710890293121338,
      "learning_rate": 0.0001999961938169475,
      "loss": 1.353,
      "step": 171
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.13916975259780884,
      "learning_rate": 0.00019999563066131124,
      "loss": 1.3254,
      "step": 172
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.11424775421619415,
      "learning_rate": 0.00019999502866852425,
      "loss": 1.3354,
      "step": 173
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1887788623571396,
      "learning_rate": 0.0001999943878388204,
      "loss": 1.2146,
      "step": 174
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.11461260914802551,
      "learning_rate": 0.00019999370817244853,
      "loss": 1.1917,
      "step": 175
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.14760863780975342,
      "learning_rate": 0.00019999298966967265,
      "loss": 1.3188,
      "step": 176
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.12025574594736099,
      "learning_rate": 0.00019999223233077177,
      "loss": 1.1911,
      "step": 177
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.10971354693174362,
      "learning_rate": 0.0001999914361560401,
      "loss": 1.2897,
      "step": 178
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.11415617167949677,
      "learning_rate": 0.00019999060114578684,
      "loss": 1.2109,
      "step": 179
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.16903188824653625,
      "learning_rate": 0.00019998972730033622,
      "loss": 1.2581,
      "step": 180
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.11982069164514542,
      "learning_rate": 0.00019998881462002778,
      "loss": 1.132,
      "step": 181
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12455673515796661,
      "learning_rate": 0.00019998786310521585,
      "loss": 1.1973,
      "step": 182
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.11733265221118927,
      "learning_rate": 0.00019998687275627006,
      "loss": 1.2331,
      "step": 183
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1810842901468277,
      "learning_rate": 0.00019998584357357502,
      "loss": 1.392,
      "step": 184
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12266435474157333,
      "learning_rate": 0.00019998477555753055,
      "loss": 1.2015,
      "step": 185
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1284114271402359,
      "learning_rate": 0.00019998366870855133,
      "loss": 1.1354,
      "step": 186
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.10606873035430908,
      "learning_rate": 0.0001999825230270673,
      "loss": 1.2233,
      "step": 187
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.10276695340871811,
      "learning_rate": 0.0001999813385135234,
      "loss": 1.2355,
      "step": 188
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.13926969468593597,
      "learning_rate": 0.00019998011516837974,
      "loss": 1.2132,
      "step": 189
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12444811314344406,
      "learning_rate": 0.0001999788529921114,
      "loss": 1.3039,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1343580186367035,
      "learning_rate": 0.0001999775519852086,
      "loss": 1.3048,
      "step": 191
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.13577254116535187,
      "learning_rate": 0.00019997621214817667,
      "loss": 1.2767,
      "step": 192
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.19905692338943481,
      "learning_rate": 0.000199974833481536,
      "loss": 1.1594,
      "step": 193
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.16784021258354187,
      "learning_rate": 0.00019997341598582195,
      "loss": 1.3171,
      "step": 194
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.17275585234165192,
      "learning_rate": 0.00019997195966158518,
      "loss": 1.2495,
      "step": 195
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12162386626005173,
      "learning_rate": 0.0001999704645093912,
      "loss": 1.229,
      "step": 196
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.10646237432956696,
      "learning_rate": 0.0001999689305298208,
      "loss": 1.2107,
      "step": 197
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12849372625350952,
      "learning_rate": 0.00019996735772346972,
      "loss": 1.2184,
      "step": 198
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.11418499797582626,
      "learning_rate": 0.00019996574609094884,
      "loss": 1.1025,
      "step": 199
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12790577113628387,
      "learning_rate": 0.00019996409563288406,
      "loss": 1.2544,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12924174964427948,
      "learning_rate": 0.00019996240634991642,
      "loss": 1.2264,
      "step": 201
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.13757866621017456,
      "learning_rate": 0.00019996067824270205,
      "loss": 1.2573,
      "step": 202
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12320806086063385,
      "learning_rate": 0.00019995891131191205,
      "loss": 1.2307,
      "step": 203
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1522417962551117,
      "learning_rate": 0.00019995710555823276,
      "loss": 1.1926,
      "step": 204
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.16503117978572845,
      "learning_rate": 0.00019995526098236547,
      "loss": 1.2454,
      "step": 205
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1482779085636139,
      "learning_rate": 0.0001999533775850266,
      "loss": 1.2759,
      "step": 206
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.15314412117004395,
      "learning_rate": 0.00019995145536694762,
      "loss": 1.2137,
      "step": 207
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12079205363988876,
      "learning_rate": 0.00019994949432887514,
      "loss": 1.2031,
      "step": 208
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.11811251193284988,
      "learning_rate": 0.00019994749447157077,
      "loss": 1.189,
      "step": 209
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1345469206571579,
      "learning_rate": 0.00019994545579581123,
      "loss": 1.1606,
      "step": 210
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1385456919670105,
      "learning_rate": 0.00019994337830238834,
      "loss": 1.2761,
      "step": 211
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12509815394878387,
      "learning_rate": 0.00019994126199210897,
      "loss": 1.1846,
      "step": 212
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1227896437048912,
      "learning_rate": 0.00019993910686579507,
      "loss": 1.1492,
      "step": 213
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12155856937170029,
      "learning_rate": 0.00019993691292428365,
      "loss": 1.2215,
      "step": 214
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.18981783092021942,
      "learning_rate": 0.00019993468016842682,
      "loss": 1.2596,
      "step": 215
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1068006306886673,
      "learning_rate": 0.00019993240859909176,
      "loss": 1.1829,
      "step": 216
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.13875330984592438,
      "learning_rate": 0.00019993009821716074,
      "loss": 1.2887,
      "step": 217
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.19742344319820404,
      "learning_rate": 0.00019992774902353105,
      "loss": 1.3287,
      "step": 218
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.17027704417705536,
      "learning_rate": 0.0001999253610191151,
      "loss": 1.2095,
      "step": 219
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.19514943659305573,
      "learning_rate": 0.00019992293420484039,
      "loss": 1.2722,
      "step": 220
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.18672405183315277,
      "learning_rate": 0.00019992046858164944,
      "loss": 1.2715,
      "step": 221
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.13758531212806702,
      "learning_rate": 0.0001999179641504999,
      "loss": 1.2543,
      "step": 222
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.13648605346679688,
      "learning_rate": 0.00019991542091236437,
      "loss": 1.212,
      "step": 223
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.19048669934272766,
      "learning_rate": 0.00019991283886823074,
      "loss": 1.255,
      "step": 224
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12030428647994995,
      "learning_rate": 0.00019991021801910177,
      "loss": 1.1612,
      "step": 225
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.10078694671392441,
      "learning_rate": 0.00019990755836599538,
      "loss": 1.1917,
      "step": 226
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.1158280149102211,
      "learning_rate": 0.0001999048599099446,
      "loss": 1.2159,
      "step": 227
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.16041213274002075,
      "learning_rate": 0.00019990212265199738,
      "loss": 1.2418,
      "step": 228
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.13681334257125854,
      "learning_rate": 0.0001998993465932169,
      "loss": 1.2027,
      "step": 229
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.14342515170574188,
      "learning_rate": 0.00019989653173468135,
      "loss": 1.2697,
      "step": 230
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12977394461631775,
      "learning_rate": 0.000199893678077484,
      "loss": 1.267,
      "step": 231
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.12351156771183014,
      "learning_rate": 0.00019989078562273314,
      "loss": 1.2558,
      "step": 232
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.10182546079158783,
      "learning_rate": 0.0001998878543715522,
      "loss": 1.1067,
      "step": 233
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.17594227194786072,
      "learning_rate": 0.0001998848843250796,
      "loss": 1.1997,
      "step": 234
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11048205941915512,
      "learning_rate": 0.00019988187548446894,
      "loss": 1.3169,
      "step": 235
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11089014261960983,
      "learning_rate": 0.0001998788278508888,
      "loss": 1.2793,
      "step": 236
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11526766419410706,
      "learning_rate": 0.00019987574142552275,
      "loss": 1.2264,
      "step": 237
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12847495079040527,
      "learning_rate": 0.00019987261620956964,
      "loss": 1.1928,
      "step": 238
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.14403772354125977,
      "learning_rate": 0.00019986945220424324,
      "loss": 1.2262,
      "step": 239
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12136774510145187,
      "learning_rate": 0.00019986624941077238,
      "loss": 1.2406,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.14794844388961792,
      "learning_rate": 0.000199863007830401,
      "loss": 1.1733,
      "step": 241
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.16121861338615417,
      "learning_rate": 0.0001998597274643881,
      "loss": 1.1862,
      "step": 242
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12165330350399017,
      "learning_rate": 0.00019985640831400776,
      "loss": 1.2149,
      "step": 243
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.14631958305835724,
      "learning_rate": 0.0001998530503805491,
      "loss": 1.1806,
      "step": 244
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12301331758499146,
      "learning_rate": 0.00019984965366531623,
      "loss": 1.1993,
      "step": 245
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.10062460601329803,
      "learning_rate": 0.00019984621816962844,
      "loss": 1.1379,
      "step": 246
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.14659780263900757,
      "learning_rate": 0.00019984274389482005,
      "loss": 1.1902,
      "step": 247
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.14320142567157745,
      "learning_rate": 0.00019983923084224045,
      "loss": 1.2294,
      "step": 248
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11687632650136948,
      "learning_rate": 0.00019983567901325403,
      "loss": 1.2552,
      "step": 249
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.16291531920433044,
      "learning_rate": 0.00019983208840924026,
      "loss": 1.2043,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1354690045118332,
      "learning_rate": 0.0001998284590315937,
      "loss": 1.3202,
      "step": 251
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.09634935110807419,
      "learning_rate": 0.00019982479088172405,
      "loss": 1.1809,
      "step": 252
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12565192580223083,
      "learning_rate": 0.00019982108396105584,
      "loss": 1.1979,
      "step": 253
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1394387185573578,
      "learning_rate": 0.00019981733827102884,
      "loss": 1.2099,
      "step": 254
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11999369412660599,
      "learning_rate": 0.00019981355381309789,
      "loss": 1.2258,
      "step": 255
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12794126570224762,
      "learning_rate": 0.0001998097305887328,
      "loss": 1.1917,
      "step": 256
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13503679633140564,
      "learning_rate": 0.00019980586859941847,
      "loss": 1.2256,
      "step": 257
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13524261116981506,
      "learning_rate": 0.00019980196784665478,
      "loss": 1.3306,
      "step": 258
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11370112746953964,
      "learning_rate": 0.00019979802833195682,
      "loss": 1.2725,
      "step": 259
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13050656020641327,
      "learning_rate": 0.00019979405005685465,
      "loss": 1.2208,
      "step": 260
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12767919898033142,
      "learning_rate": 0.00019979003302289335,
      "loss": 1.2296,
      "step": 261
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13849209249019623,
      "learning_rate": 0.0001997859772316331,
      "loss": 1.1895,
      "step": 262
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1287163943052292,
      "learning_rate": 0.00019978188268464912,
      "loss": 1.1911,
      "step": 263
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.23364481329917908,
      "learning_rate": 0.0001997777493835317,
      "loss": 1.2314,
      "step": 264
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11647038906812668,
      "learning_rate": 0.00019977357732988614,
      "loss": 1.2791,
      "step": 265
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13404060900211334,
      "learning_rate": 0.0001997693665253329,
      "loss": 1.2163,
      "step": 266
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.16389964520931244,
      "learning_rate": 0.0001997651169715073,
      "loss": 1.2097,
      "step": 267
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12130267173051834,
      "learning_rate": 0.00019976082867005984,
      "loss": 1.208,
      "step": 268
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.137754425406456,
      "learning_rate": 0.00019975650162265608,
      "loss": 1.2381,
      "step": 269
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11582165211439133,
      "learning_rate": 0.0001997521358309766,
      "loss": 1.2098,
      "step": 270
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1380978375673294,
      "learning_rate": 0.00019974773129671701,
      "loss": 1.2234,
      "step": 271
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.160974383354187,
      "learning_rate": 0.00019974328802158797,
      "loss": 1.2393,
      "step": 272
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13412973284721375,
      "learning_rate": 0.0001997388060073152,
      "loss": 1.2559,
      "step": 273
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13221485912799835,
      "learning_rate": 0.00019973428525563947,
      "loss": 1.1567,
      "step": 274
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.14068099856376648,
      "learning_rate": 0.00019972972576831656,
      "loss": 1.2943,
      "step": 275
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11903106421232224,
      "learning_rate": 0.0001997251275471174,
      "loss": 1.1296,
      "step": 276
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13461224734783173,
      "learning_rate": 0.00019972049059382782,
      "loss": 1.2962,
      "step": 277
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12724487483501434,
      "learning_rate": 0.00019971581491024873,
      "loss": 1.1718,
      "step": 278
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13860218226909637,
      "learning_rate": 0.0001997111004981962,
      "loss": 1.2252,
      "step": 279
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.19351427257061005,
      "learning_rate": 0.00019970634735950115,
      "loss": 1.2356,
      "step": 280
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.15713943541049957,
      "learning_rate": 0.00019970155549600978,
      "loss": 1.1903,
      "step": 281
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.14985191822052002,
      "learning_rate": 0.00019969672490958304,
      "loss": 1.2847,
      "step": 282
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.11453903466463089,
      "learning_rate": 0.0001996918556020972,
      "loss": 1.1793,
      "step": 283
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.10715116560459137,
      "learning_rate": 0.0001996869475754434,
      "loss": 1.1892,
      "step": 284
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.15101438760757446,
      "learning_rate": 0.00019968200083152782,
      "loss": 1.2309,
      "step": 285
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1928005963563919,
      "learning_rate": 0.00019967701537227175,
      "loss": 1.1671,
      "step": 286
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.15050560235977173,
      "learning_rate": 0.00019967199119961152,
      "loss": 1.18,
      "step": 287
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1496017426252365,
      "learning_rate": 0.0001996669283154984,
      "loss": 1.2428,
      "step": 288
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.13969042897224426,
      "learning_rate": 0.0001996618267218988,
      "loss": 1.1796,
      "step": 289
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.13866157829761505,
      "learning_rate": 0.00019965668642079408,
      "loss": 1.2416,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1374632716178894,
      "learning_rate": 0.00019965150741418073,
      "loss": 1.1569,
      "step": 291
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1628558188676834,
      "learning_rate": 0.0001996462897040702,
      "loss": 1.1786,
      "step": 292
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.13050654530525208,
      "learning_rate": 0.0001996410332924889,
      "loss": 1.2329,
      "step": 293
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.13198009133338928,
      "learning_rate": 0.0001996357381814785,
      "loss": 1.1741,
      "step": 294
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1294049471616745,
      "learning_rate": 0.00019963040437309549,
      "loss": 1.2092,
      "step": 295
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1288776695728302,
      "learning_rate": 0.00019962503186941142,
      "loss": 1.1835,
      "step": 296
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2188815176486969,
      "learning_rate": 0.00019961962067251298,
      "loss": 1.2379,
      "step": 297
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.11882001161575317,
      "learning_rate": 0.00019961417078450178,
      "loss": 1.2123,
      "step": 298
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1323470175266266,
      "learning_rate": 0.00019960868220749448,
      "loss": 1.1979,
      "step": 299
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12562941014766693,
      "learning_rate": 0.00019960315494362284,
      "loss": 1.2495,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1273299902677536,
      "learning_rate": 0.00019959758899503353,
      "loss": 1.1731,
      "step": 301
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.08701295405626297,
      "learning_rate": 0.0001995919843638883,
      "loss": 1.2509,
      "step": 302
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.11925175786018372,
      "learning_rate": 0.00019958634105236395,
      "loss": 1.1741,
      "step": 303
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1286870688199997,
      "learning_rate": 0.00019958065906265228,
      "loss": 1.2128,
      "step": 304
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.13018850982189178,
      "learning_rate": 0.0001995749383969601,
      "loss": 1.1956,
      "step": 305
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.11639239639043808,
      "learning_rate": 0.00019956917905750924,
      "loss": 1.1247,
      "step": 306
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.14018474519252777,
      "learning_rate": 0.00019956338104653657,
      "loss": 1.169,
      "step": 307
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12825161218643188,
      "learning_rate": 0.00019955754436629399,
      "loss": 1.2347,
      "step": 308
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.14431771636009216,
      "learning_rate": 0.00019955166901904837,
      "loss": 1.1933,
      "step": 309
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12925191223621368,
      "learning_rate": 0.00019954575500708162,
      "loss": 1.1441,
      "step": 310
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.15757229924201965,
      "learning_rate": 0.0001995398023326907,
      "loss": 1.1917,
      "step": 311
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.11902903020381927,
      "learning_rate": 0.00019953381099818755,
      "loss": 1.2676,
      "step": 312
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.11153929680585861,
      "learning_rate": 0.00019952778100589913,
      "loss": 1.2106,
      "step": 313
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1239159107208252,
      "learning_rate": 0.00019952171235816747,
      "loss": 1.2289,
      "step": 314
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1490047723054886,
      "learning_rate": 0.00019951560505734945,
      "loss": 1.0946,
      "step": 315
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1218082383275032,
      "learning_rate": 0.00019950945910581717,
      "loss": 1.2123,
      "step": 316
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.15356531739234924,
      "learning_rate": 0.00019950327450595764,
      "loss": 1.216,
      "step": 317
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1204916387796402,
      "learning_rate": 0.00019949705126017287,
      "loss": 1.1746,
      "step": 318
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1525362730026245,
      "learning_rate": 0.00019949078937087986,
      "loss": 1.2499,
      "step": 319
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.13765856623649597,
      "learning_rate": 0.0001994844888405107,
      "loss": 1.1274,
      "step": 320
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12787377834320068,
      "learning_rate": 0.00019947814967151244,
      "loss": 1.1931,
      "step": 321
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.18042221665382385,
      "learning_rate": 0.00019947177186634715,
      "loss": 1.2644,
      "step": 322
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12798437476158142,
      "learning_rate": 0.00019946535542749184,
      "loss": 1.2727,
      "step": 323
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.10068286210298538,
      "learning_rate": 0.00019945890035743866,
      "loss": 1.1264,
      "step": 324
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12313905358314514,
      "learning_rate": 0.00019945240665869465,
      "loss": 1.1375,
      "step": 325
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1544978767633438,
      "learning_rate": 0.00019944587433378186,
      "loss": 1.1778,
      "step": 326
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12463398277759552,
      "learning_rate": 0.0001994393033852374,
      "loss": 1.2864,
      "step": 327
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12185288965702057,
      "learning_rate": 0.00019943269381561334,
      "loss": 1.2328,
      "step": 328
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.11738427728414536,
      "learning_rate": 0.00019942604562747678,
      "loss": 1.1828,
      "step": 329
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.10502393543720245,
      "learning_rate": 0.00019941935882340976,
      "loss": 1.1813,
      "step": 330
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1403665393590927,
      "learning_rate": 0.00019941263340600939,
      "loss": 1.1451,
      "step": 331
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.15555088222026825,
      "learning_rate": 0.00019940586937788776,
      "loss": 1.2493,
      "step": 332
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.15555334091186523,
      "learning_rate": 0.0001993990667416719,
      "loss": 1.2508,
      "step": 333
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12217622250318527,
      "learning_rate": 0.0001993922255000039,
      "loss": 1.1728,
      "step": 334
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.10914736986160278,
      "learning_rate": 0.0001993853456555408,
      "loss": 1.1778,
      "step": 335
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.16819725930690765,
      "learning_rate": 0.00019937842721095468,
      "loss": 1.2392,
      "step": 336
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.17794717848300934,
      "learning_rate": 0.00019937147016893257,
      "loss": 1.2649,
      "step": 337
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.12918175756931305,
      "learning_rate": 0.00019936447453217646,
      "loss": 1.1474,
      "step": 338
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.12213893234729767,
      "learning_rate": 0.00019935744030340346,
      "loss": 1.2235,
      "step": 339
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1194637343287468,
      "learning_rate": 0.00019935036748534553,
      "loss": 1.1602,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.13917727768421173,
      "learning_rate": 0.0001993432560807497,
      "loss": 1.291,
      "step": 341
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11624502390623093,
      "learning_rate": 0.00019933610609237793,
      "loss": 1.212,
      "step": 342
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.14727282524108887,
      "learning_rate": 0.00019932891752300717,
      "loss": 1.2396,
      "step": 343
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11135949194431305,
      "learning_rate": 0.00019932169037542946,
      "loss": 1.2861,
      "step": 344
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.12972313165664673,
      "learning_rate": 0.00019931442465245165,
      "loss": 1.2686,
      "step": 345
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.15232208371162415,
      "learning_rate": 0.00019930712035689575,
      "loss": 1.1409,
      "step": 346
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.09613309800624847,
      "learning_rate": 0.00019929977749159859,
      "loss": 1.1036,
      "step": 347
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.19230255484580994,
      "learning_rate": 0.00019929239605941208,
      "loss": 1.265,
      "step": 348
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11372523009777069,
      "learning_rate": 0.0001992849760632031,
      "loss": 1.2545,
      "step": 349
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1853352189064026,
      "learning_rate": 0.00019927751750585347,
      "loss": 1.2068,
      "step": 350
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.13008828461170197,
      "learning_rate": 0.00019927002039026002,
      "loss": 1.1856,
      "step": 351
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1508476436138153,
      "learning_rate": 0.00019926248471933454,
      "loss": 1.2015,
      "step": 352
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1302882730960846,
      "learning_rate": 0.0001992549104960038,
      "loss": 1.0942,
      "step": 353
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.12283176183700562,
      "learning_rate": 0.0001992472977232095,
      "loss": 1.267,
      "step": 354
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1643277257680893,
      "learning_rate": 0.00019923964640390843,
      "loss": 1.2379,
      "step": 355
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.12542961537837982,
      "learning_rate": 0.00019923195654107225,
      "loss": 1.1525,
      "step": 356
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11767968535423279,
      "learning_rate": 0.00019922422813768758,
      "loss": 1.2277,
      "step": 357
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.14810173213481903,
      "learning_rate": 0.00019921646119675605,
      "loss": 1.2547,
      "step": 358
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1391076147556305,
      "learning_rate": 0.00019920865572129425,
      "loss": 1.1316,
      "step": 359
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11483366787433624,
      "learning_rate": 0.00019920081171433379,
      "loss": 1.2079,
      "step": 360
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.10360131412744522,
      "learning_rate": 0.00019919292917892112,
      "loss": 1.0661,
      "step": 361
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1497877538204193,
      "learning_rate": 0.00019918500811811778,
      "loss": 1.1717,
      "step": 362
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.13599057495594025,
      "learning_rate": 0.00019917704853500016,
      "loss": 1.2622,
      "step": 363
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1493876427412033,
      "learning_rate": 0.00019916905043265972,
      "loss": 1.1707,
      "step": 364
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1103268712759018,
      "learning_rate": 0.00019916101381420285,
      "loss": 1.171,
      "step": 365
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.14008721709251404,
      "learning_rate": 0.00019915293868275083,
      "loss": 1.1704,
      "step": 366
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.14860564470291138,
      "learning_rate": 0.00019914482504143995,
      "loss": 1.2032,
      "step": 367
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.12669701874256134,
      "learning_rate": 0.00019913667289342147,
      "loss": 1.1837,
      "step": 368
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11132178455591202,
      "learning_rate": 0.0001991284822418616,
      "loss": 1.2078,
      "step": 369
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.10884418338537216,
      "learning_rate": 0.00019912025308994148,
      "loss": 1.1126,
      "step": 370
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11686211079359055,
      "learning_rate": 0.00019911198544085722,
      "loss": 1.2231,
      "step": 371
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1474391222000122,
      "learning_rate": 0.00019910367929781988,
      "loss": 1.3108,
      "step": 372
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.16809217631816864,
      "learning_rate": 0.00019909533466405546,
      "loss": 1.1535,
      "step": 373
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1155829057097435,
      "learning_rate": 0.00019908695154280496,
      "loss": 1.2856,
      "step": 374
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.12333803623914719,
      "learning_rate": 0.00019907852993732424,
      "loss": 1.173,
      "step": 375
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.13569971919059753,
      "learning_rate": 0.0001990700698508842,
      "loss": 1.1998,
      "step": 376
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.10038352757692337,
      "learning_rate": 0.0001990615712867706,
      "loss": 1.223,
      "step": 377
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1252184957265854,
      "learning_rate": 0.00019905303424828417,
      "loss": 1.1854,
      "step": 378
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1042080968618393,
      "learning_rate": 0.00019904445873874068,
      "loss": 1.2585,
      "step": 379
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.09970517456531525,
      "learning_rate": 0.00019903584476147065,
      "loss": 1.1106,
      "step": 380
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.16583873331546783,
      "learning_rate": 0.00019902719231981974,
      "loss": 1.1244,
      "step": 381
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.14304110407829285,
      "learning_rate": 0.00019901850141714841,
      "loss": 1.1668,
      "step": 382
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.12279763072729111,
      "learning_rate": 0.0001990097720568321,
      "loss": 1.1734,
      "step": 383
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.18480980396270752,
      "learning_rate": 0.00019900100424226125,
      "loss": 1.2025,
      "step": 384
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.1288074553012848,
      "learning_rate": 0.00019899219797684113,
      "loss": 1.1686,
      "step": 385
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.16159610450267792,
      "learning_rate": 0.000198983353263992,
      "loss": 1.0858,
      "step": 386
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11314205825328827,
      "learning_rate": 0.00019897447010714905,
      "loss": 1.216,
      "step": 387
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.10722334682941437,
      "learning_rate": 0.00019896554850976238,
      "loss": 1.1284,
      "step": 388
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.11441335827112198,
      "learning_rate": 0.00019895658847529708,
      "loss": 1.171,
      "step": 389
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1485300213098526,
      "learning_rate": 0.00019894759000723306,
      "loss": 1.2303,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.15200264751911163,
      "learning_rate": 0.00019893855310906526,
      "loss": 1.1625,
      "step": 391
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.19079971313476562,
      "learning_rate": 0.0001989294777843035,
      "loss": 1.1611,
      "step": 392
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10569038987159729,
      "learning_rate": 0.00019892036403647254,
      "loss": 1.1826,
      "step": 393
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.12180844694375992,
      "learning_rate": 0.00019891121186911206,
      "loss": 1.1533,
      "step": 394
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.12385908514261246,
      "learning_rate": 0.00019890202128577662,
      "loss": 1.0944,
      "step": 395
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14997483789920807,
      "learning_rate": 0.00019889279229003576,
      "loss": 1.1612,
      "step": 396
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.11354799568653107,
      "learning_rate": 0.00019888352488547394,
      "loss": 1.2031,
      "step": 397
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.11700699478387833,
      "learning_rate": 0.00019887421907569048,
      "loss": 1.1516,
      "step": 398
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.11281560361385345,
      "learning_rate": 0.00019886487486429964,
      "loss": 1.1448,
      "step": 399
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.13167984783649445,
      "learning_rate": 0.00019885549225493064,
      "loss": 1.2766,
      "step": 400
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1406400203704834,
      "learning_rate": 0.00019884607125122755,
      "loss": 1.2293,
      "step": 401
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.12406296283006668,
      "learning_rate": 0.0001988366118568494,
      "loss": 1.2535,
      "step": 402
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.17210279405117035,
      "learning_rate": 0.0001988271140754701,
      "loss": 1.169,
      "step": 403
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10846318304538727,
      "learning_rate": 0.00019881757791077845,
      "loss": 1.1285,
      "step": 404
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.12754079699516296,
      "learning_rate": 0.00019880800336647824,
      "loss": 1.2149,
      "step": 405
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10800088942050934,
      "learning_rate": 0.0001987983904462881,
      "loss": 1.1799,
      "step": 406
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1004258394241333,
      "learning_rate": 0.00019878873915394154,
      "loss": 1.1755,
      "step": 407
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14554853737354279,
      "learning_rate": 0.00019877904949318703,
      "loss": 1.0679,
      "step": 408
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.15368486940860748,
      "learning_rate": 0.00019876932146778794,
      "loss": 1.1504,
      "step": 409
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14515992999076843,
      "learning_rate": 0.00019875955508152253,
      "loss": 1.1747,
      "step": 410
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.11832825094461441,
      "learning_rate": 0.0001987497503381839,
      "loss": 1.1853,
      "step": 411
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.13041706383228302,
      "learning_rate": 0.00019873990724158014,
      "loss": 1.1791,
      "step": 412
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1311672031879425,
      "learning_rate": 0.00019873002579553418,
      "loss": 1.2095,
      "step": 413
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10732489079236984,
      "learning_rate": 0.00019872010600388392,
      "loss": 1.1581,
      "step": 414
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.12021815031766891,
      "learning_rate": 0.00019871014787048197,
      "loss": 1.2195,
      "step": 415
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14570243656635284,
      "learning_rate": 0.00019870015139919606,
      "loss": 1.2547,
      "step": 416
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1345817595720291,
      "learning_rate": 0.00019869011659390866,
      "loss": 1.1773,
      "step": 417
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14794768393039703,
      "learning_rate": 0.00019868004345851716,
      "loss": 1.2125,
      "step": 418
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.13665053248405457,
      "learning_rate": 0.00019866993199693392,
      "loss": 1.1788,
      "step": 419
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10995712876319885,
      "learning_rate": 0.000198659782213086,
      "loss": 1.1014,
      "step": 420
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14954274892807007,
      "learning_rate": 0.00019864959411091556,
      "loss": 1.2366,
      "step": 421
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.12720710039138794,
      "learning_rate": 0.00019863936769437955,
      "loss": 1.1753,
      "step": 422
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1346060335636139,
      "learning_rate": 0.00019862910296744967,
      "loss": 1.186,
      "step": 423
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.13823828101158142,
      "learning_rate": 0.00019861879993411275,
      "loss": 1.196,
      "step": 424
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.16172631084918976,
      "learning_rate": 0.00019860845859837032,
      "loss": 1.2495,
      "step": 425
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10828535258769989,
      "learning_rate": 0.00019859807896423882,
      "loss": 1.1359,
      "step": 426
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1006685271859169,
      "learning_rate": 0.0001985876610357496,
      "loss": 1.211,
      "step": 427
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14297699928283691,
      "learning_rate": 0.00019857720481694885,
      "loss": 1.1523,
      "step": 428
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.15648308396339417,
      "learning_rate": 0.00019856671031189766,
      "loss": 1.211,
      "step": 429
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.13714708387851715,
      "learning_rate": 0.000198556177524672,
      "loss": 1.1308,
      "step": 430
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.11879761517047882,
      "learning_rate": 0.0001985456064593626,
      "loss": 1.2578,
      "step": 431
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.11080913990736008,
      "learning_rate": 0.00019853499712007522,
      "loss": 1.1154,
      "step": 432
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14348849654197693,
      "learning_rate": 0.00019852434951093034,
      "loss": 1.2466,
      "step": 433
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1826973557472229,
      "learning_rate": 0.00019851366363606346,
      "loss": 1.2316,
      "step": 434
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10660386830568314,
      "learning_rate": 0.00019850293949962478,
      "loss": 1.1855,
      "step": 435
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14928336441516876,
      "learning_rate": 0.00019849217710577946,
      "loss": 1.218,
      "step": 436
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1420520842075348,
      "learning_rate": 0.00019848137645870747,
      "loss": 1.157,
      "step": 437
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14897502958774567,
      "learning_rate": 0.0001984705375626036,
      "loss": 1.2023,
      "step": 438
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.14458344876766205,
      "learning_rate": 0.0001984596604216777,
      "loss": 1.1471,
      "step": 439
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.15127678215503693,
      "learning_rate": 0.0001984487450401542,
      "loss": 1.1559,
      "step": 440
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.168124258518219,
      "learning_rate": 0.00019843779142227256,
      "loss": 1.2276,
      "step": 441
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.12663902342319489,
      "learning_rate": 0.00019842679957228704,
      "loss": 1.1921,
      "step": 442
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1376810222864151,
      "learning_rate": 0.00019841576949446675,
      "loss": 1.1402,
      "step": 443
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.15001225471496582,
      "learning_rate": 0.0001984047011930956,
      "loss": 1.2595,
      "step": 444
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10642759501934052,
      "learning_rate": 0.00019839359467247242,
      "loss": 1.1276,
      "step": 445
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.13107293844223022,
      "learning_rate": 0.0001983824499369109,
      "loss": 1.1504,
      "step": 446
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14816614985466003,
      "learning_rate": 0.00019837126699073947,
      "loss": 1.1381,
      "step": 447
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.11539189517498016,
      "learning_rate": 0.00019836004583830146,
      "loss": 1.1744,
      "step": 448
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14386339485645294,
      "learning_rate": 0.00019834878648395505,
      "loss": 1.1511,
      "step": 449
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.15595972537994385,
      "learning_rate": 0.00019833748893207325,
      "loss": 1.1811,
      "step": 450
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.13360093533992767,
      "learning_rate": 0.00019832615318704389,
      "loss": 1.1976,
      "step": 451
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.11677359789609909,
      "learning_rate": 0.00019831477925326963,
      "loss": 1.2216,
      "step": 452
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.3452812135219574,
      "learning_rate": 0.00019830336713516799,
      "loss": 1.1692,
      "step": 453
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10470034927129745,
      "learning_rate": 0.00019829191683717133,
      "loss": 1.166,
      "step": 454
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.18163242936134338,
      "learning_rate": 0.00019828042836372677,
      "loss": 1.2493,
      "step": 455
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1510835438966751,
      "learning_rate": 0.00019826890171929632,
      "loss": 1.1074,
      "step": 456
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14270548522472382,
      "learning_rate": 0.00019825733690835679,
      "loss": 1.3309,
      "step": 457
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1634017825126648,
      "learning_rate": 0.00019824573393539984,
      "loss": 1.2956,
      "step": 458
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.16399599611759186,
      "learning_rate": 0.0001982340928049319,
      "loss": 1.1892,
      "step": 459
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14624950289726257,
      "learning_rate": 0.00019822241352147427,
      "loss": 1.1973,
      "step": 460
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.12140946090221405,
      "learning_rate": 0.00019821069608956307,
      "loss": 1.1846,
      "step": 461
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.13678646087646484,
      "learning_rate": 0.00019819894051374915,
      "loss": 1.135,
      "step": 462
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.13089792430400848,
      "learning_rate": 0.0001981871467985983,
      "loss": 1.2414,
      "step": 463
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.15480473637580872,
      "learning_rate": 0.00019817531494869105,
      "loss": 1.154,
      "step": 464
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1411118507385254,
      "learning_rate": 0.00019816344496862272,
      "loss": 1.1896,
      "step": 465
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1636473685503006,
      "learning_rate": 0.00019815153686300352,
      "loss": 1.1029,
      "step": 466
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14010286331176758,
      "learning_rate": 0.0001981395906364584,
      "loss": 1.1184,
      "step": 467
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.11033454537391663,
      "learning_rate": 0.00019812760629362716,
      "loss": 1.1198,
      "step": 468
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.15972144901752472,
      "learning_rate": 0.0001981155838391643,
      "loss": 1.182,
      "step": 469
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.13386288285255432,
      "learning_rate": 0.00019810352327773935,
      "loss": 1.236,
      "step": 470
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10958616435527802,
      "learning_rate": 0.00019809142461403633,
      "loss": 1.2,
      "step": 471
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1863756626844406,
      "learning_rate": 0.00019807928785275434,
      "loss": 1.1773,
      "step": 472
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.11885905265808105,
      "learning_rate": 0.0001980671129986071,
      "loss": 1.1683,
      "step": 473
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.17800796031951904,
      "learning_rate": 0.0001980549000563232,
      "loss": 1.1368,
      "step": 474
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.102497398853302,
      "learning_rate": 0.000198042649030646,
      "loss": 1.1525,
      "step": 475
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.16952811181545258,
      "learning_rate": 0.00019803035992633366,
      "loss": 1.1763,
      "step": 476
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1202457845211029,
      "learning_rate": 0.00019801803274815917,
      "loss": 1.189,
      "step": 477
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.12368443608283997,
      "learning_rate": 0.00019800566750091016,
      "loss": 1.1378,
      "step": 478
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.17240174114704132,
      "learning_rate": 0.00019799326418938924,
      "loss": 1.1983,
      "step": 479
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.10996685177087784,
      "learning_rate": 0.0001979808228184137,
      "loss": 1.2112,
      "step": 480
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.13451705873012543,
      "learning_rate": 0.0001979683433928156,
      "loss": 1.2004,
      "step": 481
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.13685095310211182,
      "learning_rate": 0.0001979558259174418,
      "loss": 1.2038,
      "step": 482
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14584603905677795,
      "learning_rate": 0.00019794327039715395,
      "loss": 1.1791,
      "step": 483
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.13103631138801575,
      "learning_rate": 0.0001979306768368285,
      "loss": 1.2314,
      "step": 484
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.12953795492649078,
      "learning_rate": 0.0001979180452413566,
      "loss": 1.1966,
      "step": 485
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.156694233417511,
      "learning_rate": 0.00019790537561564428,
      "loss": 1.2268,
      "step": 486
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1260581761598587,
      "learning_rate": 0.00019789266796461222,
      "loss": 1.0753,
      "step": 487
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14129190146923065,
      "learning_rate": 0.00019787992229319592,
      "loss": 1.4021,
      "step": 488
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14336788654327393,
      "learning_rate": 0.00019786713860634567,
      "loss": 1.1671,
      "step": 489
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1247689500451088,
      "learning_rate": 0.00019785431690902652,
      "loss": 1.2301,
      "step": 490
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.13287052512168884,
      "learning_rate": 0.00019784145720621826,
      "loss": 1.1866,
      "step": 491
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.25992584228515625,
      "learning_rate": 0.00019782855950291542,
      "loss": 1.2193,
      "step": 492
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1274813860654831,
      "learning_rate": 0.0001978156238041274,
      "loss": 1.1718,
      "step": 493
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.3104258179664612,
      "learning_rate": 0.0001978026501148782,
      "loss": 1.1185,
      "step": 494
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1215924322605133,
      "learning_rate": 0.00019778963844020665,
      "loss": 1.1295,
      "step": 495
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.14678634703159332,
      "learning_rate": 0.00019777658878516639,
      "loss": 1.1499,
      "step": 496
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.11638636887073517,
      "learning_rate": 0.0001977635011548257,
      "loss": 1.1716,
      "step": 497
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.10572238266468048,
      "learning_rate": 0.0001977503755542677,
      "loss": 1.2048,
      "step": 498
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.10839852690696716,
      "learning_rate": 0.00019773721198859022,
      "loss": 1.1589,
      "step": 499
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.14482517540454865,
      "learning_rate": 0.00019772401046290586,
      "loss": 1.2857,
      "step": 500
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13951894640922546,
      "learning_rate": 0.00019771077098234186,
      "loss": 1.1776,
      "step": 501
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.10083983838558197,
      "learning_rate": 0.00019769749355204032,
      "loss": 1.1572,
      "step": 502
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15916188061237335,
      "learning_rate": 0.00019768417817715809,
      "loss": 1.1126,
      "step": 503
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1552194058895111,
      "learning_rate": 0.00019767082486286665,
      "loss": 1.0931,
      "step": 504
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13083401322364807,
      "learning_rate": 0.0001976574336143523,
      "loss": 1.1522,
      "step": 505
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15971675515174866,
      "learning_rate": 0.00019764400443681606,
      "loss": 1.1677,
      "step": 506
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13527174293994904,
      "learning_rate": 0.00019763053733547366,
      "loss": 1.1197,
      "step": 507
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15000160038471222,
      "learning_rate": 0.0001976170323155555,
      "loss": 1.1684,
      "step": 508
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.16487297415733337,
      "learning_rate": 0.0001976034893823069,
      "loss": 1.2398,
      "step": 509
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.14279711246490479,
      "learning_rate": 0.0001975899085409876,
      "loss": 1.1954,
      "step": 510
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13949663937091827,
      "learning_rate": 0.00019757628979687246,
      "loss": 1.2014,
      "step": 511
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.201833575963974,
      "learning_rate": 0.0001975626331552507,
      "loss": 1.1337,
      "step": 512
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1114555075764656,
      "learning_rate": 0.00019754893862142643,
      "loss": 1.0798,
      "step": 513
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.12569895386695862,
      "learning_rate": 0.00019753520620071843,
      "loss": 1.2186,
      "step": 514
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1763680875301361,
      "learning_rate": 0.0001975214358984603,
      "loss": 1.1865,
      "step": 515
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15584023296833038,
      "learning_rate": 0.00019750762772000014,
      "loss": 1.0867,
      "step": 516
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1303495168685913,
      "learning_rate": 0.000197493781670701,
      "loss": 1.2204,
      "step": 517
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.16805875301361084,
      "learning_rate": 0.00019747989775594044,
      "loss": 1.1962,
      "step": 518
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13386984169483185,
      "learning_rate": 0.0001974659759811109,
      "loss": 1.2381,
      "step": 519
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.10286135971546173,
      "learning_rate": 0.00019745201635161936,
      "loss": 1.1166,
      "step": 520
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13404947519302368,
      "learning_rate": 0.00019743801887288763,
      "loss": 1.1368,
      "step": 521
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1415138840675354,
      "learning_rate": 0.0001974239835503521,
      "loss": 1.1796,
      "step": 522
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.14284133911132812,
      "learning_rate": 0.00019740991038946404,
      "loss": 1.3066,
      "step": 523
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13862961530685425,
      "learning_rate": 0.0001973957993956892,
      "loss": 1.2155,
      "step": 524
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13555672764778137,
      "learning_rate": 0.00019738165057450816,
      "loss": 1.1775,
      "step": 525
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13014444708824158,
      "learning_rate": 0.00019736746393141617,
      "loss": 1.193,
      "step": 526
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.14756646752357483,
      "learning_rate": 0.00019735323947192316,
      "loss": 1.1859,
      "step": 527
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15065009891986847,
      "learning_rate": 0.00019733897720155375,
      "loss": 1.1165,
      "step": 528
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.11419864743947983,
      "learning_rate": 0.00019732467712584722,
      "loss": 1.0802,
      "step": 529
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.12378380447626114,
      "learning_rate": 0.0001973103392503576,
      "loss": 1.1818,
      "step": 530
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15446363389492035,
      "learning_rate": 0.00019729596358065345,
      "loss": 1.1284,
      "step": 531
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15008074045181274,
      "learning_rate": 0.00019728155012231825,
      "loss": 1.1939,
      "step": 532
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.13040664792060852,
      "learning_rate": 0.00019726709888094992,
      "loss": 1.1502,
      "step": 533
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.14954447746276855,
      "learning_rate": 0.0001972526098621612,
      "loss": 1.0818,
      "step": 534
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1308390200138092,
      "learning_rate": 0.00019723808307157948,
      "loss": 1.157,
      "step": 535
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.14830364286899567,
      "learning_rate": 0.00019722351851484676,
      "loss": 1.1949,
      "step": 536
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.12840574979782104,
      "learning_rate": 0.00019720891619761974,
      "loss": 1.1674,
      "step": 537
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.14295043051242828,
      "learning_rate": 0.0001971942761255698,
      "loss": 1.2572,
      "step": 538
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.11269776523113251,
      "learning_rate": 0.00019717959830438302,
      "loss": 1.2162,
      "step": 539
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.14106760919094086,
      "learning_rate": 0.00019716488273976003,
      "loss": 1.1588,
      "step": 540
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.15790636837482452,
      "learning_rate": 0.0001971501294374162,
      "loss": 1.1764,
      "step": 541
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.1307525485754013,
      "learning_rate": 0.00019713533840308157,
      "loss": 1.2766,
      "step": 542
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.11389176547527313,
      "learning_rate": 0.00019712050964250082,
      "loss": 1.1957,
      "step": 543
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.17872031033039093,
      "learning_rate": 0.00019710564316143323,
      "loss": 1.1752,
      "step": 544
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.11800789833068848,
      "learning_rate": 0.00019709073896565275,
      "loss": 1.1702,
      "step": 545
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11185190081596375,
      "learning_rate": 0.00019707579706094807,
      "loss": 1.145,
      "step": 546
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13011296093463898,
      "learning_rate": 0.0001970608174531224,
      "loss": 1.28,
      "step": 547
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11089736968278885,
      "learning_rate": 0.0001970458001479937,
      "loss": 1.1725,
      "step": 548
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.20593015849590302,
      "learning_rate": 0.00019703074515139445,
      "loss": 1.2632,
      "step": 549
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11232755333185196,
      "learning_rate": 0.00019701565246917183,
      "loss": 1.1334,
      "step": 550
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.14979544281959534,
      "learning_rate": 0.00019700052210718777,
      "loss": 1.1529,
      "step": 551
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11698514968156815,
      "learning_rate": 0.00019698535407131862,
      "loss": 1.0737,
      "step": 552
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1430852860212326,
      "learning_rate": 0.00019697014836745553,
      "loss": 1.2084,
      "step": 553
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1309593915939331,
      "learning_rate": 0.00019695490500150418,
      "loss": 1.1259,
      "step": 554
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.14151610434055328,
      "learning_rate": 0.00019693962397938496,
      "loss": 1.1874,
      "step": 555
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.15488272905349731,
      "learning_rate": 0.00019692430530703282,
      "loss": 1.1826,
      "step": 556
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.16296882927417755,
      "learning_rate": 0.00019690894899039734,
      "loss": 1.1312,
      "step": 557
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.14506793022155762,
      "learning_rate": 0.00019689355503544275,
      "loss": 1.2317,
      "step": 558
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11540514975786209,
      "learning_rate": 0.0001968781234481479,
      "loss": 1.2252,
      "step": 559
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.14292728900909424,
      "learning_rate": 0.00019686265423450624,
      "loss": 1.1277,
      "step": 560
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1794070154428482,
      "learning_rate": 0.00019684714740052583,
      "loss": 1.1575,
      "step": 561
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.15566787123680115,
      "learning_rate": 0.00019683160295222934,
      "loss": 1.1228,
      "step": 562
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1378752589225769,
      "learning_rate": 0.00019681602089565402,
      "loss": 1.2057,
      "step": 563
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.14548130333423615,
      "learning_rate": 0.0001968004012368518,
      "loss": 1.1405,
      "step": 564
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13984113931655884,
      "learning_rate": 0.0001967847439818892,
      "loss": 1.1636,
      "step": 565
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.18710649013519287,
      "learning_rate": 0.00019676904913684727,
      "loss": 1.1605,
      "step": 566
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13264413177967072,
      "learning_rate": 0.0001967533167078217,
      "loss": 1.243,
      "step": 567
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13541360199451447,
      "learning_rate": 0.00019673754670092284,
      "loss": 1.2209,
      "step": 568
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.14854291081428528,
      "learning_rate": 0.00019672173912227553,
      "loss": 1.1805,
      "step": 569
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13455435633659363,
      "learning_rate": 0.0001967058939780193,
      "loss": 1.1445,
      "step": 570
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1006288006901741,
      "learning_rate": 0.00019669001127430816,
      "loss": 1.1009,
      "step": 571
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11872299760580063,
      "learning_rate": 0.00019667409101731083,
      "loss": 1.1426,
      "step": 572
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.12261448800563812,
      "learning_rate": 0.0001966581332132105,
      "loss": 1.1713,
      "step": 573
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11328387260437012,
      "learning_rate": 0.00019664213786820502,
      "loss": 1.1847,
      "step": 574
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.15137681365013123,
      "learning_rate": 0.00019662610498850683,
      "loss": 1.1911,
      "step": 575
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13348551094532013,
      "learning_rate": 0.00019661003458034285,
      "loss": 1.1822,
      "step": 576
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.14462800323963165,
      "learning_rate": 0.0001965939266499547,
      "loss": 1.2277,
      "step": 577
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1062132939696312,
      "learning_rate": 0.00019657778120359847,
      "loss": 1.1245,
      "step": 578
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.16391655802726746,
      "learning_rate": 0.0001965615982475449,
      "loss": 1.1131,
      "step": 579
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.12994594871997833,
      "learning_rate": 0.00019654537778807923,
      "loss": 1.0881,
      "step": 580
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11739472299814224,
      "learning_rate": 0.00019652911983150136,
      "loss": 1.2674,
      "step": 581
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.09764236956834793,
      "learning_rate": 0.0001965128243841256,
      "loss": 1.2136,
      "step": 582
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.10236222296953201,
      "learning_rate": 0.00019649649145228102,
      "loss": 1.1586,
      "step": 583
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.12298668920993805,
      "learning_rate": 0.00019648012104231106,
      "loss": 1.2033,
      "step": 584
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1523047238588333,
      "learning_rate": 0.00019646371316057383,
      "loss": 1.1322,
      "step": 585
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.10819891095161438,
      "learning_rate": 0.00019644726781344195,
      "loss": 1.1417,
      "step": 586
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11278940737247467,
      "learning_rate": 0.0001964307850073026,
      "loss": 1.2006,
      "step": 587
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13801714777946472,
      "learning_rate": 0.00019641426474855758,
      "loss": 1.087,
      "step": 588
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1324082762002945,
      "learning_rate": 0.00019639770704362307,
      "loss": 1.1613,
      "step": 589
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.13910622894763947,
      "learning_rate": 0.00019638111189892993,
      "loss": 1.2274,
      "step": 590
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11037793010473251,
      "learning_rate": 0.00019636447932092353,
      "loss": 1.1603,
      "step": 591
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.16569583117961884,
      "learning_rate": 0.0001963478093160638,
      "loss": 1.2618,
      "step": 592
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11734496802091599,
      "learning_rate": 0.00019633110189082512,
      "loss": 1.1489,
      "step": 593
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.1150767058134079,
      "learning_rate": 0.0001963143570516965,
      "loss": 1.1563,
      "step": 594
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.11039029061794281,
      "learning_rate": 0.00019629757480518143,
      "loss": 1.1673,
      "step": 595
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.14390051364898682,
      "learning_rate": 0.00019628075515779796,
      "loss": 1.2298,
      "step": 596
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.17280730605125427,
      "learning_rate": 0.0001962638981160786,
      "loss": 1.1499,
      "step": 597
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1407255232334137,
      "learning_rate": 0.00019624700368657045,
      "loss": 1.16,
      "step": 598
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14962980151176453,
      "learning_rate": 0.00019623007187583515,
      "loss": 1.2009,
      "step": 599
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14948570728302002,
      "learning_rate": 0.0001962131026904488,
      "loss": 1.0663,
      "step": 600
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1055075004696846,
      "learning_rate": 0.000196196096137002,
      "loss": 1.0857,
      "step": 601
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14880751073360443,
      "learning_rate": 0.00019617905222209996,
      "loss": 1.1391,
      "step": 602
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14764171838760376,
      "learning_rate": 0.00019616197095236227,
      "loss": 1.2701,
      "step": 603
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1394694745540619,
      "learning_rate": 0.00019614485233442316,
      "loss": 1.1582,
      "step": 604
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13042683899402618,
      "learning_rate": 0.00019612769637493126,
      "loss": 1.1205,
      "step": 605
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1441597044467926,
      "learning_rate": 0.0001961105030805498,
      "loss": 1.1972,
      "step": 606
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.10334447771310806,
      "learning_rate": 0.00019609327245795642,
      "loss": 1.1181,
      "step": 607
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.17239807546138763,
      "learning_rate": 0.00019607600451384326,
      "loss": 1.1592,
      "step": 608
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.20697452127933502,
      "learning_rate": 0.00019605869925491706,
      "loss": 1.1455,
      "step": 609
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.11728253960609436,
      "learning_rate": 0.00019604135668789896,
      "loss": 1.1175,
      "step": 610
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13436220586299896,
      "learning_rate": 0.00019602397681952463,
      "loss": 1.2133,
      "step": 611
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14829038083553314,
      "learning_rate": 0.00019600655965654412,
      "loss": 1.2141,
      "step": 612
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13354744017124176,
      "learning_rate": 0.00019598910520572218,
      "loss": 1.1272,
      "step": 613
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12864045798778534,
      "learning_rate": 0.00019597161347383784,
      "loss": 1.2507,
      "step": 614
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13885991275310516,
      "learning_rate": 0.00019595408446768472,
      "loss": 1.218,
      "step": 615
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14450395107269287,
      "learning_rate": 0.00019593651819407084,
      "loss": 1.2279,
      "step": 616
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.15366467833518982,
      "learning_rate": 0.0001959189146598188,
      "loss": 1.1884,
      "step": 617
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.11567937582731247,
      "learning_rate": 0.00019590127387176555,
      "loss": 1.1698,
      "step": 618
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1127915307879448,
      "learning_rate": 0.00019588359583676263,
      "loss": 1.0925,
      "step": 619
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12594377994537354,
      "learning_rate": 0.00019586588056167594,
      "loss": 1.1655,
      "step": 620
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14878587424755096,
      "learning_rate": 0.0001958481280533859,
      "loss": 1.1515,
      "step": 621
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.16674849390983582,
      "learning_rate": 0.0001958303383187874,
      "loss": 1.3209,
      "step": 622
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14358431100845337,
      "learning_rate": 0.00019581251136478972,
      "loss": 1.1083,
      "step": 623
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.11093273758888245,
      "learning_rate": 0.00019579464719831667,
      "loss": 1.2159,
      "step": 624
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13105937838554382,
      "learning_rate": 0.00019577674582630652,
      "loss": 1.2736,
      "step": 625
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12985555827617645,
      "learning_rate": 0.0001957588072557119,
      "loss": 1.177,
      "step": 626
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.11330024152994156,
      "learning_rate": 0.0001957408314935,
      "loss": 1.1113,
      "step": 627
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.19557195901870728,
      "learning_rate": 0.00019572281854665234,
      "loss": 1.1992,
      "step": 628
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12178636342287064,
      "learning_rate": 0.00019570476842216498,
      "loss": 1.183,
      "step": 629
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12888066470623016,
      "learning_rate": 0.00019568668112704838,
      "loss": 1.0751,
      "step": 630
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1707497537136078,
      "learning_rate": 0.00019566855666832743,
      "loss": 1.2405,
      "step": 631
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.15435902774333954,
      "learning_rate": 0.00019565039505304145,
      "loss": 1.1772,
      "step": 632
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12436584383249283,
      "learning_rate": 0.0001956321962882442,
      "loss": 1.1606,
      "step": 633
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13563527166843414,
      "learning_rate": 0.0001956139603810039,
      "loss": 1.1612,
      "step": 634
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.16036967933177948,
      "learning_rate": 0.00019559568733840314,
      "loss": 1.27,
      "step": 635
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13167190551757812,
      "learning_rate": 0.00019557737716753896,
      "loss": 1.1621,
      "step": 636
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12179706245660782,
      "learning_rate": 0.00019555902987552283,
      "loss": 1.1819,
      "step": 637
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.12835131585597992,
      "learning_rate": 0.00019554064546948064,
      "loss": 1.1873,
      "step": 638
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14116138219833374,
      "learning_rate": 0.0001955222239565526,
      "loss": 1.2465,
      "step": 639
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.11441957950592041,
      "learning_rate": 0.00019550376534389357,
      "loss": 1.2427,
      "step": 640
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1230890154838562,
      "learning_rate": 0.00019548526963867252,
      "loss": 1.1276,
      "step": 641
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.10991024971008301,
      "learning_rate": 0.00019546673684807302,
      "loss": 1.187,
      "step": 642
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13264533877372742,
      "learning_rate": 0.000195448166979293,
      "loss": 1.1394,
      "step": 643
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14079582691192627,
      "learning_rate": 0.00019542956003954477,
      "loss": 1.1796,
      "step": 644
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13179342448711395,
      "learning_rate": 0.00019541091603605506,
      "loss": 1.1669,
      "step": 645
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.13341064751148224,
      "learning_rate": 0.000195392234976065,
      "loss": 1.2268,
      "step": 646
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.14082780480384827,
      "learning_rate": 0.00019537351686683003,
      "loss": 1.1905,
      "step": 647
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.21938788890838623,
      "learning_rate": 0.00019535476171562012,
      "loss": 1.1655,
      "step": 648
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.16186971962451935,
      "learning_rate": 0.00019533596952971954,
      "loss": 1.1718,
      "step": 649
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.13709618151187897,
      "learning_rate": 0.00019531714031642696,
      "loss": 1.1806,
      "step": 650
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12352917343378067,
      "learning_rate": 0.0001952982740830554,
      "loss": 1.1663,
      "step": 651
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.14980009198188782,
      "learning_rate": 0.00019527937083693231,
      "loss": 1.1482,
      "step": 652
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.11314062029123306,
      "learning_rate": 0.0001952604305853995,
      "loss": 1.1928,
      "step": 653
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1183028370141983,
      "learning_rate": 0.00019524145333581317,
      "loss": 1.1976,
      "step": 654
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.15007442235946655,
      "learning_rate": 0.00019522243909554377,
      "loss": 1.1954,
      "step": 655
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.13498777151107788,
      "learning_rate": 0.00019520338787197629,
      "loss": 1.2062,
      "step": 656
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1379648745059967,
      "learning_rate": 0.00019518429967251,
      "loss": 1.1986,
      "step": 657
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.11804603040218353,
      "learning_rate": 0.00019516517450455853,
      "loss": 1.2603,
      "step": 658
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.14057467877864838,
      "learning_rate": 0.00019514601237554988,
      "loss": 1.2265,
      "step": 659
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.11606588959693909,
      "learning_rate": 0.00019512681329292636,
      "loss": 1.1844,
      "step": 660
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.13914859294891357,
      "learning_rate": 0.00019510757726414472,
      "loss": 1.1636,
      "step": 661
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1327597051858902,
      "learning_rate": 0.000195088304296676,
      "loss": 1.1408,
      "step": 662
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.13434965908527374,
      "learning_rate": 0.00019506899439800557,
      "loss": 1.1862,
      "step": 663
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1302722692489624,
      "learning_rate": 0.0001950496475756332,
      "loss": 1.1566,
      "step": 664
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.16586624085903168,
      "learning_rate": 0.000195030263837073,
      "loss": 1.1307,
      "step": 665
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1201547235250473,
      "learning_rate": 0.00019501084318985335,
      "loss": 1.1717,
      "step": 666
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.14068834483623505,
      "learning_rate": 0.000194991385641517,
      "loss": 1.2196,
      "step": 667
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12063725292682648,
      "learning_rate": 0.00019497189119962105,
      "loss": 1.2287,
      "step": 668
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12668149173259735,
      "learning_rate": 0.00019495235987173693,
      "loss": 1.2103,
      "step": 669
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.10452651977539062,
      "learning_rate": 0.00019493279166545038,
      "loss": 1.1536,
      "step": 670
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.15395420789718628,
      "learning_rate": 0.00019491318658836142,
      "loss": 1.1773,
      "step": 671
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.11785230040550232,
      "learning_rate": 0.0001948935446480845,
      "loss": 1.2184,
      "step": 672
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1097661480307579,
      "learning_rate": 0.0001948738658522483,
      "loss": 1.1698,
      "step": 673
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.14201129972934723,
      "learning_rate": 0.00019485415020849582,
      "loss": 1.1408,
      "step": 674
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12449251115322113,
      "learning_rate": 0.0001948343977244844,
      "loss": 1.1461,
      "step": 675
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1279807984828949,
      "learning_rate": 0.00019481460840788573,
      "loss": 1.1351,
      "step": 676
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.15456727147102356,
      "learning_rate": 0.00019479478226638565,
      "loss": 1.2178,
      "step": 677
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.15588179230690002,
      "learning_rate": 0.0001947749193076845,
      "loss": 1.1529,
      "step": 678
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.13936276733875275,
      "learning_rate": 0.00019475501953949672,
      "loss": 1.0889,
      "step": 679
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1005239263176918,
      "learning_rate": 0.00019473508296955126,
      "loss": 1.2045,
      "step": 680
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.13752202689647675,
      "learning_rate": 0.0001947151096055912,
      "loss": 1.3045,
      "step": 681
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.14124339818954468,
      "learning_rate": 0.00019469509945537397,
      "loss": 1.2676,
      "step": 682
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12712977826595306,
      "learning_rate": 0.00019467505252667127,
      "loss": 1.1565,
      "step": 683
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1461494415998459,
      "learning_rate": 0.0001946549688272691,
      "loss": 1.2025,
      "step": 684
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.11743483692407608,
      "learning_rate": 0.0001946348483649678,
      "loss": 1.1522,
      "step": 685
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.16727003455162048,
      "learning_rate": 0.0001946146911475818,
      "loss": 1.2301,
      "step": 686
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.11091189086437225,
      "learning_rate": 0.00019459449718294008,
      "loss": 1.2669,
      "step": 687
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1614570915699005,
      "learning_rate": 0.00019457426647888563,
      "loss": 1.1449,
      "step": 688
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1451413929462433,
      "learning_rate": 0.00019455399904327585,
      "loss": 1.1697,
      "step": 689
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1327025443315506,
      "learning_rate": 0.00019453369488398238,
      "loss": 1.1493,
      "step": 690
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.14180000126361847,
      "learning_rate": 0.00019451335400889116,
      "loss": 1.1823,
      "step": 691
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.15521658957004547,
      "learning_rate": 0.00019449297642590228,
      "loss": 1.157,
      "step": 692
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12727421522140503,
      "learning_rate": 0.00019447256214293025,
      "loss": 1.2284,
      "step": 693
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1328200399875641,
      "learning_rate": 0.00019445211116790362,
      "loss": 1.1313,
      "step": 694
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.13853183388710022,
      "learning_rate": 0.00019443162350876546,
      "loss": 1.2031,
      "step": 695
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.09956386685371399,
      "learning_rate": 0.0001944110991734728,
      "loss": 1.1614,
      "step": 696
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12095305323600769,
      "learning_rate": 0.00019439053816999716,
      "loss": 1.0835,
      "step": 697
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.12095881253480911,
      "learning_rate": 0.00019436994050632414,
      "loss": 1.1573,
      "step": 698
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.18181845545768738,
      "learning_rate": 0.00019434930619045368,
      "loss": 1.1321,
      "step": 699
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.1470692902803421,
      "learning_rate": 0.00019432863523039987,
      "loss": 1.1482,
      "step": 700
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.09635978192090988,
      "learning_rate": 0.00019430792763419107,
      "loss": 1.1626,
      "step": 701
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.10199765861034393,
      "learning_rate": 0.00019428718340986988,
      "loss": 1.1828,
      "step": 702
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.13784775137901306,
      "learning_rate": 0.00019426640256549313,
      "loss": 1.1598,
      "step": 703
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12078046798706055,
      "learning_rate": 0.00019424558510913186,
      "loss": 1.2003,
      "step": 704
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.15991659462451935,
      "learning_rate": 0.00019422473104887134,
      "loss": 1.2331,
      "step": 705
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11827164143323898,
      "learning_rate": 0.000194203840392811,
      "loss": 1.1217,
      "step": 706
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.15089261531829834,
      "learning_rate": 0.00019418291314906457,
      "loss": 1.1483,
      "step": 707
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12755881249904633,
      "learning_rate": 0.00019416194932576,
      "loss": 1.2621,
      "step": 708
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.14577150344848633,
      "learning_rate": 0.00019414094893103928,
      "loss": 1.1991,
      "step": 709
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12625181674957275,
      "learning_rate": 0.00019411991197305879,
      "loss": 1.1384,
      "step": 710
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12226023524999619,
      "learning_rate": 0.00019409883845998904,
      "loss": 1.1485,
      "step": 711
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1401190161705017,
      "learning_rate": 0.00019407772840001472,
      "loss": 1.2308,
      "step": 712
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12819750607013702,
      "learning_rate": 0.00019405658180133477,
      "loss": 1.0804,
      "step": 713
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1357603371143341,
      "learning_rate": 0.00019403539867216224,
      "loss": 1.1372,
      "step": 714
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12599067389965057,
      "learning_rate": 0.00019401417902072446,
      "loss": 1.214,
      "step": 715
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.13161331415176392,
      "learning_rate": 0.00019399292285526284,
      "loss": 1.0995,
      "step": 716
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.13047954440116882,
      "learning_rate": 0.00019397163018403308,
      "loss": 1.145,
      "step": 717
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.15905041992664337,
      "learning_rate": 0.00019395030101530502,
      "loss": 1.1142,
      "step": 718
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1298377811908722,
      "learning_rate": 0.0001939289353573626,
      "loss": 1.1858,
      "step": 719
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.13681833446025848,
      "learning_rate": 0.00019390753321850404,
      "loss": 1.1442,
      "step": 720
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.14008808135986328,
      "learning_rate": 0.00019388609460704168,
      "loss": 1.259,
      "step": 721
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12642453610897064,
      "learning_rate": 0.000193864619531302,
      "loss": 1.1728,
      "step": 722
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.09796850383281708,
      "learning_rate": 0.00019384310799962573,
      "loss": 1.1347,
      "step": 723
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12592381238937378,
      "learning_rate": 0.00019382156002036764,
      "loss": 1.2043,
      "step": 724
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.14348763227462769,
      "learning_rate": 0.00019379997560189675,
      "loss": 1.1795,
      "step": 725
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12223166227340698,
      "learning_rate": 0.0001937783547525962,
      "loss": 1.1857,
      "step": 726
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.15415464341640472,
      "learning_rate": 0.00019375669748086327,
      "loss": 1.1387,
      "step": 727
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.145643949508667,
      "learning_rate": 0.00019373500379510938,
      "loss": 1.1693,
      "step": 728
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.18320897221565247,
      "learning_rate": 0.00019371327370376016,
      "loss": 1.2185,
      "step": 729
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12028311938047409,
      "learning_rate": 0.00019369150721525527,
      "loss": 1.1308,
      "step": 730
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11793339997529984,
      "learning_rate": 0.0001936697043380486,
      "loss": 1.1752,
      "step": 731
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.13715144991874695,
      "learning_rate": 0.00019364786508060808,
      "loss": 1.1919,
      "step": 732
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.17611292004585266,
      "learning_rate": 0.0001936259894514159,
      "loss": 1.1388,
      "step": 733
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12219825387001038,
      "learning_rate": 0.00019360407745896827,
      "loss": 1.277,
      "step": 734
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.15724538266658783,
      "learning_rate": 0.00019358212911177556,
      "loss": 1.3061,
      "step": 735
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1519387811422348,
      "learning_rate": 0.0001935601444183622,
      "loss": 1.1026,
      "step": 736
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.17859023809432983,
      "learning_rate": 0.0001935381233872669,
      "loss": 1.1473,
      "step": 737
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1678827702999115,
      "learning_rate": 0.00019351606602704228,
      "loss": 1.1316,
      "step": 738
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11409956216812134,
      "learning_rate": 0.0001934939723462552,
      "loss": 1.1897,
      "step": 739
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.14641015231609344,
      "learning_rate": 0.00019347184235348662,
      "loss": 1.1926,
      "step": 740
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.12848740816116333,
      "learning_rate": 0.00019344967605733153,
      "loss": 1.0354,
      "step": 741
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.10948581993579865,
      "learning_rate": 0.0001934274734663991,
      "loss": 1.208,
      "step": 742
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.09426518529653549,
      "learning_rate": 0.00019340523458931253,
      "loss": 1.1123,
      "step": 743
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1526237428188324,
      "learning_rate": 0.00019338295943470914,
      "loss": 1.1838,
      "step": 744
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.13264890015125275,
      "learning_rate": 0.00019336064801124035,
      "loss": 1.1951,
      "step": 745
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11499577760696411,
      "learning_rate": 0.0001933383003275717,
      "loss": 1.2626,
      "step": 746
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11725181341171265,
      "learning_rate": 0.0001933159163923827,
      "loss": 1.1071,
      "step": 747
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11343241482973099,
      "learning_rate": 0.00019329349621436708,
      "loss": 1.1422,
      "step": 748
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11493702977895737,
      "learning_rate": 0.00019327103980223254,
      "loss": 1.2195,
      "step": 749
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.15254105627536774,
      "learning_rate": 0.0001932485471647009,
      "loss": 1.1977,
      "step": 750
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11614365875720978,
      "learning_rate": 0.00019322601831050804,
      "loss": 1.1799,
      "step": 751
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11599418520927429,
      "learning_rate": 0.00019320345324840395,
      "loss": 1.15,
      "step": 752
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.15112927556037903,
      "learning_rate": 0.00019318085198715256,
      "loss": 1.1525,
      "step": 753
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1360999196767807,
      "learning_rate": 0.000193158214535532,
      "loss": 1.2263,
      "step": 754
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.15088684856891632,
      "learning_rate": 0.00019313554090233436,
      "loss": 1.1729,
      "step": 755
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.12696442008018494,
      "learning_rate": 0.00019311283109636583,
      "loss": 1.1787,
      "step": 756
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1401941329240799,
      "learning_rate": 0.00019309008512644667,
      "loss": 1.1748,
      "step": 757
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.15695013105869293,
      "learning_rate": 0.0001930673030014111,
      "loss": 1.0967,
      "step": 758
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1442236751317978,
      "learning_rate": 0.00019304448473010747,
      "loss": 1.1603,
      "step": 759
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1404504030942917,
      "learning_rate": 0.00019302163032139814,
      "loss": 1.1156,
      "step": 760
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11967029422521591,
      "learning_rate": 0.00019299873978415947,
      "loss": 1.0795,
      "step": 761
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.15531378984451294,
      "learning_rate": 0.00019297581312728186,
      "loss": 1.1652,
      "step": 762
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.13321006298065186,
      "learning_rate": 0.0001929528503596698,
      "loss": 1.2609,
      "step": 763
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11637604236602783,
      "learning_rate": 0.00019292985149024178,
      "loss": 1.1451,
      "step": 764
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14264105260372162,
      "learning_rate": 0.00019290681652793027,
      "loss": 1.1525,
      "step": 765
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.10428521782159805,
      "learning_rate": 0.0001928837454816818,
      "loss": 1.1316,
      "step": 766
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.15706567466259003,
      "learning_rate": 0.00019286063836045685,
      "loss": 1.2195,
      "step": 767
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11076734215021133,
      "learning_rate": 0.00019283749517322999,
      "loss": 1.2193,
      "step": 768
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.12966276705265045,
      "learning_rate": 0.00019281431592898978,
      "loss": 1.2351,
      "step": 769
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.10885287821292877,
      "learning_rate": 0.0001927911006367388,
      "loss": 1.2047,
      "step": 770
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.13525056838989258,
      "learning_rate": 0.0001927678493054935,
      "loss": 1.1472,
      "step": 771
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1555631309747696,
      "learning_rate": 0.00019274456194428454,
      "loss": 1.1545,
      "step": 772
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.13756494224071503,
      "learning_rate": 0.0001927212385621564,
      "loss": 1.1418,
      "step": 773
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.13009966909885406,
      "learning_rate": 0.00019269787916816763,
      "loss": 1.1851,
      "step": 774
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.15855556726455688,
      "learning_rate": 0.00019267448377139075,
      "loss": 1.0987,
      "step": 775
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.12294095009565353,
      "learning_rate": 0.00019265105238091228,
      "loss": 1.0513,
      "step": 776
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11121544986963272,
      "learning_rate": 0.00019262758500583263,
      "loss": 1.1689,
      "step": 777
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14975811541080475,
      "learning_rate": 0.00019260408165526637,
      "loss": 1.1498,
      "step": 778
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1533942073583603,
      "learning_rate": 0.00019258054233834183,
      "loss": 1.1847,
      "step": 779
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14544305205345154,
      "learning_rate": 0.00019255696706420148,
      "loss": 1.231,
      "step": 780
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1170126274228096,
      "learning_rate": 0.00019253335584200164,
      "loss": 1.1066,
      "step": 781
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14521895349025726,
      "learning_rate": 0.00019250970868091267,
      "loss": 1.1996,
      "step": 782
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.10920170694589615,
      "learning_rate": 0.00019248602559011882,
      "loss": 1.1863,
      "step": 783
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11857956647872925,
      "learning_rate": 0.00019246230657881834,
      "loss": 1.1909,
      "step": 784
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11749962717294693,
      "learning_rate": 0.00019243855165622343,
      "loss": 1.1982,
      "step": 785
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14940260350704193,
      "learning_rate": 0.00019241476083156026,
      "loss": 1.2241,
      "step": 786
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11718334257602692,
      "learning_rate": 0.00019239093411406886,
      "loss": 1.1309,
      "step": 787
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14892031252384186,
      "learning_rate": 0.00019236707151300327,
      "loss": 1.1198,
      "step": 788
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.13735298812389374,
      "learning_rate": 0.00019234317303763144,
      "loss": 1.1679,
      "step": 789
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11552087217569351,
      "learning_rate": 0.00019231923869723528,
      "loss": 1.1976,
      "step": 790
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14653298258781433,
      "learning_rate": 0.0001922952685011106,
      "loss": 1.2539,
      "step": 791
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.12973880767822266,
      "learning_rate": 0.00019227126245856716,
      "loss": 1.0903,
      "step": 792
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.12964066863059998,
      "learning_rate": 0.00019224722057892862,
      "loss": 1.1143,
      "step": 793
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.14856146275997162,
      "learning_rate": 0.00019222314287153255,
      "loss": 1.1265,
      "step": 794
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.16408243775367737,
      "learning_rate": 0.00019219902934573048,
      "loss": 1.2787,
      "step": 795
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.18303731083869934,
      "learning_rate": 0.00019217488001088784,
      "loss": 1.131,
      "step": 796
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1683109998703003,
      "learning_rate": 0.00019215069487638395,
      "loss": 1.2283,
      "step": 797
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.12621058523654938,
      "learning_rate": 0.000192126473951612,
      "loss": 1.1675,
      "step": 798
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.16309668123722076,
      "learning_rate": 0.0001921022172459791,
      "loss": 1.1541,
      "step": 799
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.13970138132572174,
      "learning_rate": 0.00019207792476890641,
      "loss": 1.1115,
      "step": 800
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1527499407529831,
      "learning_rate": 0.00019205359652982868,
      "loss": 1.1365,
      "step": 801
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1402420550584793,
      "learning_rate": 0.0001920292325381948,
      "loss": 1.2095,
      "step": 802
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.12534691393375397,
      "learning_rate": 0.00019200483280346748,
      "loss": 1.0911,
      "step": 803
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11527200788259506,
      "learning_rate": 0.00019198039733512326,
      "loss": 1.2426,
      "step": 804
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.11363585293292999,
      "learning_rate": 0.00019195592614265261,
      "loss": 1.1872,
      "step": 805
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13403239846229553,
      "learning_rate": 0.00019193141923555984,
      "loss": 1.1965,
      "step": 806
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.07909289747476578,
      "learning_rate": 0.00019190687662336316,
      "loss": 1.125,
      "step": 807
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.15210190415382385,
      "learning_rate": 0.00019188229831559466,
      "loss": 1.1392,
      "step": 808
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13335134088993073,
      "learning_rate": 0.00019185768432180024,
      "loss": 1.2112,
      "step": 809
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.11885562539100647,
      "learning_rate": 0.00019183303465153973,
      "loss": 1.2321,
      "step": 810
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.21687716245651245,
      "learning_rate": 0.00019180834931438673,
      "loss": 1.1389,
      "step": 811
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13110552728176117,
      "learning_rate": 0.00019178362831992878,
      "loss": 1.1144,
      "step": 812
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13845552504062653,
      "learning_rate": 0.00019175887167776717,
      "loss": 1.2079,
      "step": 813
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13883961737155914,
      "learning_rate": 0.0001917340793975172,
      "loss": 1.1783,
      "step": 814
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12044838815927505,
      "learning_rate": 0.0001917092514888078,
      "loss": 1.2217,
      "step": 815
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.2059183120727539,
      "learning_rate": 0.00019168438796128193,
      "loss": 1.1402,
      "step": 816
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.14659661054611206,
      "learning_rate": 0.0001916594888245962,
      "loss": 1.114,
      "step": 817
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12323901057243347,
      "learning_rate": 0.00019163455408842124,
      "loss": 1.1273,
      "step": 818
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.15282109379768372,
      "learning_rate": 0.00019160958376244136,
      "loss": 1.1406,
      "step": 819
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12821415066719055,
      "learning_rate": 0.00019158457785635477,
      "loss": 1.1944,
      "step": 820
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.14684921503067017,
      "learning_rate": 0.00019155953637987346,
      "loss": 1.1843,
      "step": 821
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.0899011418223381,
      "learning_rate": 0.0001915344593427233,
      "loss": 1.2003,
      "step": 822
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.11044415831565857,
      "learning_rate": 0.00019150934675464382,
      "loss": 1.1252,
      "step": 823
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12383430451154709,
      "learning_rate": 0.00019148419862538858,
      "loss": 1.2395,
      "step": 824
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.11939737945795059,
      "learning_rate": 0.00019145901496472472,
      "loss": 1.1654,
      "step": 825
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.14097991585731506,
      "learning_rate": 0.00019143379578243335,
      "loss": 1.2382,
      "step": 826
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12895333766937256,
      "learning_rate": 0.0001914085410883093,
      "loss": 1.2143,
      "step": 827
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.15892530977725983,
      "learning_rate": 0.00019138325089216118,
      "loss": 1.1687,
      "step": 828
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1418997049331665,
      "learning_rate": 0.00019135792520381142,
      "loss": 1.17,
      "step": 829
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13350597023963928,
      "learning_rate": 0.00019133256403309625,
      "loss": 1.1126,
      "step": 830
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.10527560114860535,
      "learning_rate": 0.0001913071673898656,
      "loss": 1.1397,
      "step": 831
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.11102679371833801,
      "learning_rate": 0.0001912817352839833,
      "loss": 1.2317,
      "step": 832
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13355593383312225,
      "learning_rate": 0.00019125626772532683,
      "loss": 1.1296,
      "step": 833
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1273818016052246,
      "learning_rate": 0.00019123076472378752,
      "loss": 1.1916,
      "step": 834
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13792456686496735,
      "learning_rate": 0.00019120522628927045,
      "loss": 1.1117,
      "step": 835
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1252220720052719,
      "learning_rate": 0.00019117965243169445,
      "loss": 1.0648,
      "step": 836
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12100314348936081,
      "learning_rate": 0.00019115404316099213,
      "loss": 1.1927,
      "step": 837
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.10851076990365982,
      "learning_rate": 0.00019112839848710978,
      "loss": 1.1222,
      "step": 838
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13705699145793915,
      "learning_rate": 0.00019110271842000755,
      "loss": 1.2227,
      "step": 839
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12671199440956116,
      "learning_rate": 0.00019107700296965927,
      "loss": 1.0903,
      "step": 840
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12343940138816833,
      "learning_rate": 0.0001910512521460525,
      "loss": 1.1443,
      "step": 841
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1702061891555786,
      "learning_rate": 0.00019102546595918857,
      "loss": 1.1028,
      "step": 842
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13838818669319153,
      "learning_rate": 0.00019099964441908257,
      "loss": 1.2761,
      "step": 843
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.14309564232826233,
      "learning_rate": 0.00019097378753576324,
      "loss": 1.245,
      "step": 844
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1351996809244156,
      "learning_rate": 0.00019094789531927316,
      "loss": 1.1329,
      "step": 845
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.09281519055366516,
      "learning_rate": 0.00019092196777966848,
      "loss": 1.2351,
      "step": 846
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1401689648628235,
      "learning_rate": 0.00019089600492701924,
      "loss": 1.1826,
      "step": 847
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1332876980304718,
      "learning_rate": 0.00019087000677140908,
      "loss": 1.1949,
      "step": 848
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.10782511532306671,
      "learning_rate": 0.00019084397332293538,
      "loss": 1.1521,
      "step": 849
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.11072120070457458,
      "learning_rate": 0.00019081790459170926,
      "loss": 1.1477,
      "step": 850
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.14709565043449402,
      "learning_rate": 0.00019079180058785547,
      "loss": 1.1839,
      "step": 851
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.124468132853508,
      "learning_rate": 0.00019076566132151252,
      "loss": 1.1464,
      "step": 852
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.14037898182868958,
      "learning_rate": 0.0001907394868028326,
      "loss": 1.2098,
      "step": 853
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.13780300319194794,
      "learning_rate": 0.0001907132770419816,
      "loss": 1.1203,
      "step": 854
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.11857779324054718,
      "learning_rate": 0.0001906870320491391,
      "loss": 1.1737,
      "step": 855
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.20058339834213257,
      "learning_rate": 0.00019066075183449835,
      "loss": 1.1387,
      "step": 856
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.12808549404144287,
      "learning_rate": 0.00019063443640826623,
      "loss": 1.1915,
      "step": 857
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.129967600107193,
      "learning_rate": 0.0001906080857806634,
      "loss": 1.2008,
      "step": 858
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12804092466831207,
      "learning_rate": 0.00019058169996192417,
      "loss": 1.1337,
      "step": 859
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.13986673951148987,
      "learning_rate": 0.00019055527896229643,
      "loss": 1.2626,
      "step": 860
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.10740997642278671,
      "learning_rate": 0.00019052882279204178,
      "loss": 1.229,
      "step": 861
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1163017526268959,
      "learning_rate": 0.00019050233146143556,
      "loss": 1.1362,
      "step": 862
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.14085571467876434,
      "learning_rate": 0.00019047580498076662,
      "loss": 1.1148,
      "step": 863
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.10840459167957306,
      "learning_rate": 0.00019044924336033758,
      "loss": 1.1519,
      "step": 864
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.15908709168434143,
      "learning_rate": 0.0001904226466104647,
      "loss": 1.1112,
      "step": 865
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.11741416156291962,
      "learning_rate": 0.0001903960147414778,
      "loss": 1.2063,
      "step": 866
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.11850143224000931,
      "learning_rate": 0.0001903693477637204,
      "loss": 1.1535,
      "step": 867
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.11395768076181412,
      "learning_rate": 0.00019034264568754968,
      "loss": 1.1537,
      "step": 868
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1616373211145401,
      "learning_rate": 0.00019031590852333635,
      "loss": 1.2063,
      "step": 869
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.10760870575904846,
      "learning_rate": 0.00019028913628146488,
      "loss": 1.0885,
      "step": 870
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12765026092529297,
      "learning_rate": 0.00019026232897233328,
      "loss": 1.1817,
      "step": 871
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.16606953740119934,
      "learning_rate": 0.0001902354866063532,
      "loss": 1.186,
      "step": 872
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.13426344096660614,
      "learning_rate": 0.0001902086091939499,
      "loss": 1.2091,
      "step": 873
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.11393071711063385,
      "learning_rate": 0.00019018169674556227,
      "loss": 1.2361,
      "step": 874
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12144559621810913,
      "learning_rate": 0.0001901547492716428,
      "loss": 1.1635,
      "step": 875
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12127339839935303,
      "learning_rate": 0.00019012776678265756,
      "loss": 1.1595,
      "step": 876
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1189318522810936,
      "learning_rate": 0.00019010074928908622,
      "loss": 1.1726,
      "step": 877
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.15319706499576569,
      "learning_rate": 0.0001900736968014221,
      "loss": 1.1569,
      "step": 878
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12083688378334045,
      "learning_rate": 0.0001900466093301721,
      "loss": 1.2283,
      "step": 879
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.11533675342798233,
      "learning_rate": 0.0001900194868858566,
      "loss": 1.1413,
      "step": 880
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12481196224689484,
      "learning_rate": 0.0001899923294790097,
      "loss": 1.0807,
      "step": 881
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.11788033694028854,
      "learning_rate": 0.00018996513712017898,
      "loss": 1.1547,
      "step": 882
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1465139240026474,
      "learning_rate": 0.0001899379098199257,
      "loss": 1.1614,
      "step": 883
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.10817444324493408,
      "learning_rate": 0.0001899106475888246,
      "loss": 1.1078,
      "step": 884
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.10764222592115402,
      "learning_rate": 0.00018988335043746403,
      "loss": 1.0928,
      "step": 885
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.15646237134933472,
      "learning_rate": 0.00018985601837644587,
      "loss": 1.16,
      "step": 886
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1301606297492981,
      "learning_rate": 0.00018982865141638558,
      "loss": 1.2066,
      "step": 887
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.16527211666107178,
      "learning_rate": 0.00018980124956791215,
      "loss": 1.2106,
      "step": 888
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.14493213593959808,
      "learning_rate": 0.00018977381284166817,
      "loss": 1.1443,
      "step": 889
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12763580679893494,
      "learning_rate": 0.00018974634124830977,
      "loss": 1.1442,
      "step": 890
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12920235097408295,
      "learning_rate": 0.0001897188347985066,
      "loss": 1.2153,
      "step": 891
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1180356815457344,
      "learning_rate": 0.00018969129350294178,
      "loss": 1.1542,
      "step": 892
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1473442018032074,
      "learning_rate": 0.00018966371737231207,
      "loss": 1.212,
      "step": 893
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.13634775578975677,
      "learning_rate": 0.00018963610641732777,
      "loss": 1.1645,
      "step": 894
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.15170378983020782,
      "learning_rate": 0.00018960846064871258,
      "loss": 1.2001,
      "step": 895
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.09580370783805847,
      "learning_rate": 0.00018958078007720385,
      "loss": 1.1612,
      "step": 896
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.08840633928775787,
      "learning_rate": 0.00018955306471355235,
      "loss": 1.1404,
      "step": 897
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.14565491676330566,
      "learning_rate": 0.00018952531456852247,
      "loss": 1.1083,
      "step": 898
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.10930919647216797,
      "learning_rate": 0.00018949752965289197,
      "loss": 1.222,
      "step": 899
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1713230311870575,
      "learning_rate": 0.0001894697099774523,
      "loss": 1.156,
      "step": 900
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12931941449642181,
      "learning_rate": 0.0001894418555530082,
      "loss": 1.1665,
      "step": 901
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.11677331477403641,
      "learning_rate": 0.000189413966390378,
      "loss": 1.1792,
      "step": 902
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.10117232799530029,
      "learning_rate": 0.00018938604250039363,
      "loss": 1.1828,
      "step": 903
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.12400182336568832,
      "learning_rate": 0.00018935808389390033,
      "loss": 1.1387,
      "step": 904
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.09708792716264725,
      "learning_rate": 0.0001893300905817569,
      "loss": 1.1844,
      "step": 905
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.151262566447258,
      "learning_rate": 0.00018930206257483564,
      "loss": 1.1505,
      "step": 906
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.10400430113077164,
      "learning_rate": 0.00018927399988402232,
      "loss": 1.141,
      "step": 907
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.11995876580476761,
      "learning_rate": 0.00018924590252021614,
      "loss": 1.1783,
      "step": 908
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.09842579811811447,
      "learning_rate": 0.00018921777049432984,
      "loss": 1.097,
      "step": 909
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1297786384820938,
      "learning_rate": 0.00018918960381728947,
      "loss": 1.089,
      "step": 910
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.12886491417884827,
      "learning_rate": 0.00018916140250003474,
      "loss": 1.1288,
      "step": 911
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10425812751054764,
      "learning_rate": 0.0001891331665535187,
      "loss": 1.137,
      "step": 912
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13873575627803802,
      "learning_rate": 0.00018910489598870784,
      "loss": 1.2034,
      "step": 913
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.14942969381809235,
      "learning_rate": 0.00018907659081658214,
      "loss": 1.1956,
      "step": 914
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.20343375205993652,
      "learning_rate": 0.00018904825104813498,
      "loss": 1.0601,
      "step": 915
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1407237946987152,
      "learning_rate": 0.0001890198766943732,
      "loss": 1.1543,
      "step": 916
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.17368385195732117,
      "learning_rate": 0.00018899146776631712,
      "loss": 1.099,
      "step": 917
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13458490371704102,
      "learning_rate": 0.0001889630242750004,
      "loss": 1.1479,
      "step": 918
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10542963445186615,
      "learning_rate": 0.0001889345462314702,
      "loss": 1.1118,
      "step": 919
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.12162148952484131,
      "learning_rate": 0.000188906033646787,
      "loss": 1.1273,
      "step": 920
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1052945926785469,
      "learning_rate": 0.00018887748653202477,
      "loss": 1.1246,
      "step": 921
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1347377747297287,
      "learning_rate": 0.00018884890489827096,
      "loss": 1.2216,
      "step": 922
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.11131931841373444,
      "learning_rate": 0.00018882028875662627,
      "loss": 1.1546,
      "step": 923
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.11395307630300522,
      "learning_rate": 0.0001887916381182049,
      "loss": 1.1592,
      "step": 924
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.11989081650972366,
      "learning_rate": 0.00018876295299413443,
      "loss": 1.0387,
      "step": 925
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.15784479677677155,
      "learning_rate": 0.00018873423339555584,
      "loss": 1.1356,
      "step": 926
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1340600997209549,
      "learning_rate": 0.0001887054793336235,
      "loss": 1.1998,
      "step": 927
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.12768340110778809,
      "learning_rate": 0.0001886766908195051,
      "loss": 1.1325,
      "step": 928
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10033055394887924,
      "learning_rate": 0.00018864786786438185,
      "loss": 1.1625,
      "step": 929
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.14583489298820496,
      "learning_rate": 0.00018861901047944818,
      "loss": 1.1366,
      "step": 930
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.15575261414051056,
      "learning_rate": 0.00018859011867591202,
      "loss": 1.2406,
      "step": 931
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.11372552067041397,
      "learning_rate": 0.00018856119246499458,
      "loss": 1.157,
      "step": 932
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2012961059808731,
      "learning_rate": 0.00018853223185793047,
      "loss": 1.1067,
      "step": 933
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13451699912548065,
      "learning_rate": 0.00018850323686596764,
      "loss": 1.1617,
      "step": 934
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13018018007278442,
      "learning_rate": 0.00018847420750036747,
      "loss": 1.1927,
      "step": 935
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10275855660438538,
      "learning_rate": 0.00018844514377240457,
      "loss": 1.1966,
      "step": 936
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13736280798912048,
      "learning_rate": 0.000188416045693367,
      "loss": 1.2217,
      "step": 937
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1374405473470688,
      "learning_rate": 0.0001883869132745561,
      "loss": 1.1748,
      "step": 938
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1224416196346283,
      "learning_rate": 0.0001883577465272866,
      "loss": 1.1821,
      "step": 939
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13809524476528168,
      "learning_rate": 0.00018832854546288642,
      "loss": 1.1672,
      "step": 940
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.12225989252328873,
      "learning_rate": 0.00018829931009269705,
      "loss": 1.1925,
      "step": 941
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.12609559297561646,
      "learning_rate": 0.0001882700404280731,
      "loss": 1.2357,
      "step": 942
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10144169628620148,
      "learning_rate": 0.00018824073648038257,
      "loss": 1.1657,
      "step": 943
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.11947871744632721,
      "learning_rate": 0.0001882113982610068,
      "loss": 1.0728,
      "step": 944
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13083475828170776,
      "learning_rate": 0.00018818202578134042,
      "loss": 1.1456,
      "step": 945
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.12034618109464645,
      "learning_rate": 0.0001881526190527913,
      "loss": 1.1877,
      "step": 946
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1417878121137619,
      "learning_rate": 0.00018812317808678076,
      "loss": 1.1393,
      "step": 947
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1185133084654808,
      "learning_rate": 0.0001880937028947433,
      "loss": 1.1681,
      "step": 948
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.15700748562812805,
      "learning_rate": 0.00018806419348812672,
      "loss": 1.1909,
      "step": 949
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.09864174574613571,
      "learning_rate": 0.00018803464987839216,
      "loss": 1.2071,
      "step": 950
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1282501220703125,
      "learning_rate": 0.000188005072077014,
      "loss": 1.1548,
      "step": 951
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10471979528665543,
      "learning_rate": 0.00018797546009547995,
      "loss": 1.0656,
      "step": 952
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13534162938594818,
      "learning_rate": 0.0001879458139452909,
      "loss": 1.1352,
      "step": 953
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.11762665212154388,
      "learning_rate": 0.00018791613363796118,
      "loss": 1.1249,
      "step": 954
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.14869233965873718,
      "learning_rate": 0.00018788641918501814,
      "loss": 1.1866,
      "step": 955
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13467518985271454,
      "learning_rate": 0.00018785667059800264,
      "loss": 1.182,
      "step": 956
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.13973596692085266,
      "learning_rate": 0.00018782688788846865,
      "loss": 1.125,
      "step": 957
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3526693284511566,
      "learning_rate": 0.0001877970710679834,
      "loss": 1.1822,
      "step": 958
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.11632002890110016,
      "learning_rate": 0.0001877672201481275,
      "loss": 1.0813,
      "step": 959
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10728446394205093,
      "learning_rate": 0.0001877373351404946,
      "loss": 1.2281,
      "step": 960
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.12606677412986755,
      "learning_rate": 0.00018770741605669171,
      "loss": 1.1368,
      "step": 961
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.09479843080043793,
      "learning_rate": 0.0001876774629083391,
      "loss": 1.1177,
      "step": 962
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.10453743487596512,
      "learning_rate": 0.00018764747570707019,
      "loss": 1.0443,
      "step": 963
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13196226954460144,
      "learning_rate": 0.00018761745446453168,
      "loss": 1.1934,
      "step": 964
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1571948528289795,
      "learning_rate": 0.00018758739919238349,
      "loss": 1.1536,
      "step": 965
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.11835550516843796,
      "learning_rate": 0.00018755730990229868,
      "loss": 1.1222,
      "step": 966
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1546671837568283,
      "learning_rate": 0.00018752718660596367,
      "loss": 1.158,
      "step": 967
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13039018213748932,
      "learning_rate": 0.00018749702931507796,
      "loss": 1.1316,
      "step": 968
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.12123449891805649,
      "learning_rate": 0.00018746683804135429,
      "loss": 1.2026,
      "step": 969
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13353972136974335,
      "learning_rate": 0.00018743661279651855,
      "loss": 1.1934,
      "step": 970
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.15473197400569916,
      "learning_rate": 0.00018740635359231,
      "loss": 1.2703,
      "step": 971
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1353679746389389,
      "learning_rate": 0.00018737606044048086,
      "loss": 1.1561,
      "step": 972
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13205945491790771,
      "learning_rate": 0.00018734573335279668,
      "loss": 1.1405,
      "step": 973
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.17842771112918854,
      "learning_rate": 0.00018731537234103617,
      "loss": 1.1976,
      "step": 974
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.16799959540367126,
      "learning_rate": 0.00018728497741699117,
      "loss": 1.1696,
      "step": 975
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.12035121768712997,
      "learning_rate": 0.0001872545485924667,
      "loss": 1.1202,
      "step": 976
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.17218442261219025,
      "learning_rate": 0.00018722408587928103,
      "loss": 1.1212,
      "step": 977
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.12618304789066315,
      "learning_rate": 0.00018719358928926546,
      "loss": 1.1363,
      "step": 978
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.14692512154579163,
      "learning_rate": 0.00018716305883426454,
      "loss": 1.1842,
      "step": 979
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.11900829523801804,
      "learning_rate": 0.00018713249452613598,
      "loss": 1.1348,
      "step": 980
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.14168883860111237,
      "learning_rate": 0.00018710189637675056,
      "loss": 1.1595,
      "step": 981
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1520184427499771,
      "learning_rate": 0.00018707126439799224,
      "loss": 1.2239,
      "step": 982
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.12150763720273972,
      "learning_rate": 0.00018704059860175817,
      "loss": 1.1374,
      "step": 983
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.14397048950195312,
      "learning_rate": 0.00018700989899995857,
      "loss": 1.1687,
      "step": 984
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13999448716640472,
      "learning_rate": 0.00018697916560451682,
      "loss": 1.1344,
      "step": 985
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.16015517711639404,
      "learning_rate": 0.0001869483984273694,
      "loss": 1.0954,
      "step": 986
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1412123143672943,
      "learning_rate": 0.00018691759748046594,
      "loss": 1.1372,
      "step": 987
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.12798652052879333,
      "learning_rate": 0.00018688676277576916,
      "loss": 1.1595,
      "step": 988
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13193997740745544,
      "learning_rate": 0.00018685589432525492,
      "loss": 1.1869,
      "step": 989
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.10416474938392639,
      "learning_rate": 0.0001868249921409122,
      "loss": 1.1717,
      "step": 990
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1703733205795288,
      "learning_rate": 0.00018679405623474296,
      "loss": 1.1867,
      "step": 991
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.16422846913337708,
      "learning_rate": 0.00018676308661876243,
      "loss": 1.1797,
      "step": 992
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.22485192120075226,
      "learning_rate": 0.00018673208330499883,
      "loss": 1.2193,
      "step": 993
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.14356638491153717,
      "learning_rate": 0.00018670104630549348,
      "loss": 1.1652,
      "step": 994
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.10687747597694397,
      "learning_rate": 0.0001866699756323008,
      "loss": 1.1609,
      "step": 995
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1718064844608307,
      "learning_rate": 0.0001866388712974883,
      "loss": 1.1732,
      "step": 996
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1390850991010666,
      "learning_rate": 0.0001866077333131365,
      "loss": 1.2401,
      "step": 997
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1195969358086586,
      "learning_rate": 0.00018657656169133907,
      "loss": 1.1355,
      "step": 998
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.140546515583992,
      "learning_rate": 0.00018654535644420275,
      "loss": 1.2246,
      "step": 999
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13566848635673523,
      "learning_rate": 0.00018651411758384718,
      "loss": 1.2768,
      "step": 1000
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13083189725875854,
      "learning_rate": 0.0001864828451224053,
      "loss": 1.1845,
      "step": 1001
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1323072910308838,
      "learning_rate": 0.00018645153907202285,
      "loss": 1.1456,
      "step": 1002
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1162799522280693,
      "learning_rate": 0.00018642019944485884,
      "loss": 1.1227,
      "step": 1003
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.12003195285797119,
      "learning_rate": 0.00018638882625308522,
      "loss": 1.2015,
      "step": 1004
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13853693008422852,
      "learning_rate": 0.0001863574195088869,
      "loss": 1.1261,
      "step": 1005
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1668931245803833,
      "learning_rate": 0.00018632597922446195,
      "loss": 1.2139,
      "step": 1006
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.10058322548866272,
      "learning_rate": 0.0001862945054120214,
      "loss": 1.2924,
      "step": 1007
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1276826709508896,
      "learning_rate": 0.00018626299808378934,
      "loss": 1.159,
      "step": 1008
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.116209477186203,
      "learning_rate": 0.00018623145725200278,
      "loss": 1.1276,
      "step": 1009
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13649022579193115,
      "learning_rate": 0.0001861998829289119,
      "loss": 1.1007,
      "step": 1010
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.15929585695266724,
      "learning_rate": 0.0001861682751267798,
      "loss": 1.0971,
      "step": 1011
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.14585983753204346,
      "learning_rate": 0.00018613663385788252,
      "loss": 1.1947,
      "step": 1012
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.14330320060253143,
      "learning_rate": 0.0001861049591345092,
      "loss": 1.2389,
      "step": 1013
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13158376514911652,
      "learning_rate": 0.00018607325096896195,
      "loss": 1.1084,
      "step": 1014
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11147231608629227,
      "learning_rate": 0.00018604150937355586,
      "loss": 1.1907,
      "step": 1015
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13745777308940887,
      "learning_rate": 0.00018600973436061897,
      "loss": 1.1994,
      "step": 1016
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.15902560949325562,
      "learning_rate": 0.00018597792594249234,
      "loss": 1.1217,
      "step": 1017
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12467870116233826,
      "learning_rate": 0.00018594608413153001,
      "loss": 1.1227,
      "step": 1018
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1277403086423874,
      "learning_rate": 0.00018591420894009897,
      "loss": 1.1606,
      "step": 1019
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1249706819653511,
      "learning_rate": 0.00018588230038057914,
      "loss": 1.1529,
      "step": 1020
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12876665592193604,
      "learning_rate": 0.00018585035846536348,
      "loss": 1.0964,
      "step": 1021
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13685166835784912,
      "learning_rate": 0.00018581838320685782,
      "loss": 1.135,
      "step": 1022
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12459905445575714,
      "learning_rate": 0.00018578637461748106,
      "loss": 1.1959,
      "step": 1023
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.143964946269989,
      "learning_rate": 0.00018575433270966486,
      "loss": 1.1723,
      "step": 1024
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.16013987362384796,
      "learning_rate": 0.00018572225749585402,
      "loss": 1.0888,
      "step": 1025
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12811952829360962,
      "learning_rate": 0.00018569014898850612,
      "loss": 1.1676,
      "step": 1026
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11166799068450928,
      "learning_rate": 0.0001856580072000918,
      "loss": 1.1786,
      "step": 1027
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12708692252635956,
      "learning_rate": 0.00018562583214309445,
      "loss": 1.1048,
      "step": 1028
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.17268818616867065,
      "learning_rate": 0.0001855936238300106,
      "loss": 1.1541,
      "step": 1029
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.10485857725143433,
      "learning_rate": 0.00018556138227334955,
      "loss": 1.1839,
      "step": 1030
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11118695884943008,
      "learning_rate": 0.00018552910748563357,
      "loss": 1.1792,
      "step": 1031
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13639196753501892,
      "learning_rate": 0.00018549679947939778,
      "loss": 1.217,
      "step": 1032
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11770455539226532,
      "learning_rate": 0.00018546445826719024,
      "loss": 1.2678,
      "step": 1033
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.08685890585184097,
      "learning_rate": 0.00018543208386157193,
      "loss": 1.1447,
      "step": 1034
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11796125024557114,
      "learning_rate": 0.00018539967627511668,
      "loss": 1.1642,
      "step": 1035
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.20064766705036163,
      "learning_rate": 0.00018536723552041123,
      "loss": 1.1996,
      "step": 1036
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.08910904079675674,
      "learning_rate": 0.00018533476161005518,
      "loss": 1.0655,
      "step": 1037
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11469966173171997,
      "learning_rate": 0.00018530225455666102,
      "loss": 1.0882,
      "step": 1038
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.10993573069572449,
      "learning_rate": 0.00018526971437285417,
      "loss": 1.1941,
      "step": 1039
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11953170597553253,
      "learning_rate": 0.00018523714107127276,
      "loss": 1.1806,
      "step": 1040
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12015917897224426,
      "learning_rate": 0.00018520453466456797,
      "loss": 1.1641,
      "step": 1041
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11989867687225342,
      "learning_rate": 0.00018517189516540378,
      "loss": 1.1641,
      "step": 1042
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1269235610961914,
      "learning_rate": 0.00018513922258645685,
      "loss": 1.172,
      "step": 1043
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.15240049362182617,
      "learning_rate": 0.000185106516940417,
      "loss": 1.1619,
      "step": 1044
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.10796070098876953,
      "learning_rate": 0.00018507377823998665,
      "loss": 1.107,
      "step": 1045
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.17335449159145355,
      "learning_rate": 0.0001850410064978811,
      "loss": 1.1559,
      "step": 1046
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11667191237211227,
      "learning_rate": 0.00018500820172682856,
      "loss": 1.1111,
      "step": 1047
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12778350710868835,
      "learning_rate": 0.00018497536393957003,
      "loss": 1.2216,
      "step": 1048
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13233643770217896,
      "learning_rate": 0.0001849424931488593,
      "loss": 1.0906,
      "step": 1049
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.09946810454130173,
      "learning_rate": 0.00018490958936746306,
      "loss": 1.1869,
      "step": 1050
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13828347623348236,
      "learning_rate": 0.0001848766526081607,
      "loss": 1.1845,
      "step": 1051
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1186758428812027,
      "learning_rate": 0.0001848436828837445,
      "loss": 1.1627,
      "step": 1052
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12951210141181946,
      "learning_rate": 0.00018481068020701952,
      "loss": 1.1593,
      "step": 1053
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.16224415600299835,
      "learning_rate": 0.00018477764459080362,
      "loss": 1.1376,
      "step": 1054
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.16603530943393707,
      "learning_rate": 0.00018474457604792746,
      "loss": 1.1532,
      "step": 1055
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13803525269031525,
      "learning_rate": 0.00018471147459123448,
      "loss": 1.2309,
      "step": 1056
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.15626363456249237,
      "learning_rate": 0.00018467834023358086,
      "loss": 1.1285,
      "step": 1057
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12634095549583435,
      "learning_rate": 0.0001846451729878357,
      "loss": 1.1179,
      "step": 1058
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11677097529172897,
      "learning_rate": 0.00018461197286688067,
      "loss": 1.1402,
      "step": 1059
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.15240176022052765,
      "learning_rate": 0.00018457873988361037,
      "loss": 1.1865,
      "step": 1060
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1185385137796402,
      "learning_rate": 0.00018454547405093212,
      "loss": 1.1503,
      "step": 1061
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.14304830133914948,
      "learning_rate": 0.00018451217538176598,
      "loss": 1.2013,
      "step": 1062
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.12168017774820328,
      "learning_rate": 0.0001844788438890447,
      "loss": 1.1824,
      "step": 1063
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11695750057697296,
      "learning_rate": 0.00018444547958571396,
      "loss": 1.1363,
      "step": 1064
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.11034349352121353,
      "learning_rate": 0.00018441208248473197,
      "loss": 1.136,
      "step": 1065
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.13765186071395874,
      "learning_rate": 0.00018437865259906985,
      "loss": 1.1274,
      "step": 1066
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.19514743983745575,
      "learning_rate": 0.00018434518994171135,
      "loss": 1.2204,
      "step": 1067
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11248396337032318,
      "learning_rate": 0.000184311694525653,
      "loss": 1.2519,
      "step": 1068
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.12585227191448212,
      "learning_rate": 0.00018427816636390398,
      "loss": 1.1139,
      "step": 1069
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11345751583576202,
      "learning_rate": 0.0001842446054694863,
      "loss": 1.161,
      "step": 1070
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.18978716433048248,
      "learning_rate": 0.00018421101185543465,
      "loss": 1.0791,
      "step": 1071
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.08767321705818176,
      "learning_rate": 0.0001841773855347963,
      "loss": 1.1452,
      "step": 1072
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1056608185172081,
      "learning_rate": 0.0001841437265206314,
      "loss": 1.1743,
      "step": 1073
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11408628523349762,
      "learning_rate": 0.0001841100348260127,
      "loss": 1.1023,
      "step": 1074
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.0937858298420906,
      "learning_rate": 0.0001840763104640257,
      "loss": 1.164,
      "step": 1075
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1392996460199356,
      "learning_rate": 0.0001840425534477685,
      "loss": 1.1874,
      "step": 1076
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.09266506880521774,
      "learning_rate": 0.000184008763790352,
      "loss": 1.1173,
      "step": 1077
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1418595016002655,
      "learning_rate": 0.00018397494150489965,
      "loss": 1.1442,
      "step": 1078
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.13778012990951538,
      "learning_rate": 0.00018394108660454766,
      "loss": 1.1213,
      "step": 1079
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11514226347208023,
      "learning_rate": 0.00018390719910244487,
      "loss": 1.0873,
      "step": 1080
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1019645631313324,
      "learning_rate": 0.00018387327901175284,
      "loss": 1.1713,
      "step": 1081
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.14194844663143158,
      "learning_rate": 0.00018383932634564572,
      "loss": 1.1494,
      "step": 1082
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.12397710978984833,
      "learning_rate": 0.00018380534111731028,
      "loss": 1.1547,
      "step": 1083
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11762266606092453,
      "learning_rate": 0.00018377132333994607,
      "loss": 1.2773,
      "step": 1084
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.13840502500534058,
      "learning_rate": 0.0001837372730267652,
      "loss": 1.168,
      "step": 1085
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.14093951880931854,
      "learning_rate": 0.00018370319019099234,
      "loss": 1.1236,
      "step": 1086
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.16627205908298492,
      "learning_rate": 0.00018366907484586496,
      "loss": 1.1575,
      "step": 1087
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1065717339515686,
      "learning_rate": 0.000183634927004633,
      "loss": 1.1191,
      "step": 1088
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.10662219673395157,
      "learning_rate": 0.00018360074668055914,
      "loss": 1.1314,
      "step": 1089
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1636817753314972,
      "learning_rate": 0.00018356653388691859,
      "loss": 1.1133,
      "step": 1090
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1271677017211914,
      "learning_rate": 0.00018353228863699923,
      "loss": 1.1998,
      "step": 1091
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11342174559831619,
      "learning_rate": 0.00018349801094410148,
      "loss": 1.1719,
      "step": 1092
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1364445835351944,
      "learning_rate": 0.00018346370082153844,
      "loss": 1.1398,
      "step": 1093
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.14579923450946808,
      "learning_rate": 0.00018342935828263575,
      "loss": 1.2527,
      "step": 1094
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1637580245733261,
      "learning_rate": 0.00018339498334073168,
      "loss": 1.121,
      "step": 1095
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11182312667369843,
      "learning_rate": 0.000183360576009177,
      "loss": 1.16,
      "step": 1096
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.08931689709424973,
      "learning_rate": 0.00018332613630133517,
      "loss": 1.1902,
      "step": 1097
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1516425758600235,
      "learning_rate": 0.0001832916642305822,
      "loss": 1.2429,
      "step": 1098
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.12618716061115265,
      "learning_rate": 0.00018325715981030657,
      "loss": 1.2123,
      "step": 1099
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11010251194238663,
      "learning_rate": 0.00018322262305390947,
      "loss": 1.1837,
      "step": 1100
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1476753205060959,
      "learning_rate": 0.00018318805397480453,
      "loss": 1.0707,
      "step": 1101
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1369474232196808,
      "learning_rate": 0.00018315345258641802,
      "loss": 1.17,
      "step": 1102
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.12050425261259079,
      "learning_rate": 0.00018311881890218873,
      "loss": 1.1268,
      "step": 1103
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11010442674160004,
      "learning_rate": 0.0001830841529355679,
      "loss": 1.1269,
      "step": 1104
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.13477829098701477,
      "learning_rate": 0.0001830494547000195,
      "loss": 1.0979,
      "step": 1105
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.14473913609981537,
      "learning_rate": 0.00018301472420901983,
      "loss": 1.1524,
      "step": 1106
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.13764269649982452,
      "learning_rate": 0.00018297996147605787,
      "loss": 1.172,
      "step": 1107
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11180257797241211,
      "learning_rate": 0.00018294516651463507,
      "loss": 1.0786,
      "step": 1108
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.14511972665786743,
      "learning_rate": 0.00018291033933826533,
      "loss": 1.1228,
      "step": 1109
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.11991332471370697,
      "learning_rate": 0.0001828754799604752,
      "loss": 1.1967,
      "step": 1110
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.12486451864242554,
      "learning_rate": 0.00018284058839480361,
      "loss": 1.1394,
      "step": 1111
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.12407375872135162,
      "learning_rate": 0.00018280566465480206,
      "loss": 1.1809,
      "step": 1112
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.09744204580783844,
      "learning_rate": 0.00018277070875403456,
      "loss": 1.1706,
      "step": 1113
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.18626387417316437,
      "learning_rate": 0.00018273572070607753,
      "loss": 1.2052,
      "step": 1114
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1360238790512085,
      "learning_rate": 0.00018270070052451993,
      "loss": 1.166,
      "step": 1115
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.12996245920658112,
      "learning_rate": 0.00018266564822296322,
      "loss": 1.1722,
      "step": 1116
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.13286609947681427,
      "learning_rate": 0.0001826305638150213,
      "loss": 1.1631,
      "step": 1117
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.11805126070976257,
      "learning_rate": 0.0001825954473143205,
      "loss": 1.1611,
      "step": 1118
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.12335482239723206,
      "learning_rate": 0.00018256029873449974,
      "loss": 1.132,
      "step": 1119
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13978764414787292,
      "learning_rate": 0.00018252511808921032,
      "loss": 1.1739,
      "step": 1120
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1658758819103241,
      "learning_rate": 0.00018248990539211594,
      "loss": 1.1005,
      "step": 1121
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.12452206760644913,
      "learning_rate": 0.00018245466065689282,
      "loss": 1.1628,
      "step": 1122
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13508792221546173,
      "learning_rate": 0.00018241938389722967,
      "loss": 1.1651,
      "step": 1123
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.11108937114477158,
      "learning_rate": 0.0001823840751268275,
      "loss": 1.1273,
      "step": 1124
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.12401359528303146,
      "learning_rate": 0.00018234873435939988,
      "loss": 1.1381,
      "step": 1125
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.12407370656728745,
      "learning_rate": 0.00018231336160867275,
      "loss": 1.1432,
      "step": 1126
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13364069163799286,
      "learning_rate": 0.00018227795688838445,
      "loss": 1.1129,
      "step": 1127
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13108260929584503,
      "learning_rate": 0.00018224252021228579,
      "loss": 1.3399,
      "step": 1128
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.09451694786548615,
      "learning_rate": 0.00018220705159413994,
      "loss": 1.0702,
      "step": 1129
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1250794380903244,
      "learning_rate": 0.00018217155104772258,
      "loss": 1.2317,
      "step": 1130
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1759403944015503,
      "learning_rate": 0.0001821360185868216,
      "loss": 1.0398,
      "step": 1131
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.11398587375879288,
      "learning_rate": 0.00018210045422523745,
      "loss": 1.1165,
      "step": 1132
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1343258023262024,
      "learning_rate": 0.00018206485797678295,
      "loss": 1.2181,
      "step": 1133
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.14392144978046417,
      "learning_rate": 0.00018202922985528322,
      "loss": 1.1785,
      "step": 1134
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.10283941775560379,
      "learning_rate": 0.00018199356987457586,
      "loss": 1.0988,
      "step": 1135
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.14914466440677643,
      "learning_rate": 0.00018195787804851078,
      "loss": 1.1606,
      "step": 1136
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.12123484164476395,
      "learning_rate": 0.00018192215439095024,
      "loss": 1.1301,
      "step": 1137
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13725829124450684,
      "learning_rate": 0.00018188639891576893,
      "loss": 1.1914,
      "step": 1138
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.15940611064434052,
      "learning_rate": 0.00018185061163685385,
      "loss": 1.1275,
      "step": 1139
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1224663108587265,
      "learning_rate": 0.0001818147925681044,
      "loss": 1.2581,
      "step": 1140
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.09736640006303787,
      "learning_rate": 0.00018177894172343226,
      "loss": 1.1369,
      "step": 1141
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13415399193763733,
      "learning_rate": 0.00018174305911676148,
      "loss": 1.1669,
      "step": 1142
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1012444719672203,
      "learning_rate": 0.00018170714476202848,
      "loss": 1.182,
      "step": 1143
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.15265336632728577,
      "learning_rate": 0.00018167119867318197,
      "loss": 1.1865,
      "step": 1144
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.14122100174427032,
      "learning_rate": 0.00018163522086418299,
      "loss": 1.1917,
      "step": 1145
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.10353901237249374,
      "learning_rate": 0.00018159921134900487,
      "loss": 1.1616,
      "step": 1146
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1090007871389389,
      "learning_rate": 0.00018156317014163338,
      "loss": 1.2013,
      "step": 1147
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.10143119841814041,
      "learning_rate": 0.0001815270972560664,
      "loss": 1.1629,
      "step": 1148
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13200704753398895,
      "learning_rate": 0.00018149099270631434,
      "loss": 1.1462,
      "step": 1149
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.144054114818573,
      "learning_rate": 0.00018145485650639974,
      "loss": 1.1599,
      "step": 1150
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.14410661160945892,
      "learning_rate": 0.00018141868867035745,
      "loss": 1.2169,
      "step": 1151
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.126825213432312,
      "learning_rate": 0.00018138248921223467,
      "loss": 1.1378,
      "step": 1152
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.12881851196289062,
      "learning_rate": 0.00018134625814609083,
      "loss": 1.178,
      "step": 1153
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.14542536437511444,
      "learning_rate": 0.00018130999548599767,
      "loss": 1.0958,
      "step": 1154
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1356719434261322,
      "learning_rate": 0.00018127370124603926,
      "loss": 1.1321,
      "step": 1155
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13868890702724457,
      "learning_rate": 0.00018123737544031176,
      "loss": 1.1292,
      "step": 1156
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.10863742232322693,
      "learning_rate": 0.00018120101808292372,
      "loss": 1.1482,
      "step": 1157
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.17541877925395966,
      "learning_rate": 0.000181164629187996,
      "loss": 1.1582,
      "step": 1158
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13811787962913513,
      "learning_rate": 0.00018112820876966152,
      "loss": 1.209,
      "step": 1159
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.19145584106445312,
      "learning_rate": 0.00018109175684206559,
      "loss": 1.2338,
      "step": 1160
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.11923053115606308,
      "learning_rate": 0.00018105527341936574,
      "loss": 1.1749,
      "step": 1161
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.11127019673585892,
      "learning_rate": 0.0001810187585157317,
      "loss": 1.1743,
      "step": 1162
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.10044596344232559,
      "learning_rate": 0.00018098221214534542,
      "loss": 1.1479,
      "step": 1163
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.12105502933263779,
      "learning_rate": 0.00018094563432240107,
      "loss": 1.1292,
      "step": 1164
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1040533185005188,
      "learning_rate": 0.0001809090250611051,
      "loss": 1.1555,
      "step": 1165
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.18880759179592133,
      "learning_rate": 0.00018087238437567614,
      "loss": 1.1243,
      "step": 1166
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.1332799792289734,
      "learning_rate": 0.00018083571228034496,
      "loss": 1.1422,
      "step": 1167
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.13694879412651062,
      "learning_rate": 0.0001807990087893546,
      "loss": 1.1374,
      "step": 1168
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.12513794004917145,
      "learning_rate": 0.00018076227391696027,
      "loss": 1.1672,
      "step": 1169
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12300128489732742,
      "learning_rate": 0.0001807255076774294,
      "loss": 1.204,
      "step": 1170
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12841059267520905,
      "learning_rate": 0.00018068871008504152,
      "loss": 1.1682,
      "step": 1171
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.11650604754686356,
      "learning_rate": 0.00018065188115408843,
      "loss": 1.1746,
      "step": 1172
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.15397092700004578,
      "learning_rate": 0.00018061502089887407,
      "loss": 1.1567,
      "step": 1173
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09518057852983475,
      "learning_rate": 0.0001805781293337145,
      "loss": 1.1135,
      "step": 1174
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1298142373561859,
      "learning_rate": 0.00018054120647293797,
      "loss": 1.103,
      "step": 1175
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.11090854555368423,
      "learning_rate": 0.00018050425233088495,
      "loss": 1.1852,
      "step": 1176
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.11291561275720596,
      "learning_rate": 0.000180467266921908,
      "loss": 1.1414,
      "step": 1177
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.130789652466774,
      "learning_rate": 0.00018043025026037176,
      "loss": 1.1101,
      "step": 1178
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.15158888697624207,
      "learning_rate": 0.00018039320236065313,
      "loss": 1.1865,
      "step": 1179
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1113128513097763,
      "learning_rate": 0.0001803561232371411,
      "loss": 1.1012,
      "step": 1180
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.14126601815223694,
      "learning_rate": 0.0001803190129042367,
      "loss": 1.1409,
      "step": 1181
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1358482837677002,
      "learning_rate": 0.00018028187137635325,
      "loss": 1.1401,
      "step": 1182
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.10074830055236816,
      "learning_rate": 0.000180244698667916,
      "loss": 1.1125,
      "step": 1183
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1303671896457672,
      "learning_rate": 0.00018020749479336247,
      "loss": 1.2261,
      "step": 1184
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.19165180623531342,
      "learning_rate": 0.00018017025976714218,
      "loss": 1.1587,
      "step": 1185
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12580236792564392,
      "learning_rate": 0.00018013299360371686,
      "loss": 1.0776,
      "step": 1186
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12789411842823029,
      "learning_rate": 0.00018009569631756013,
      "loss": 1.1559,
      "step": 1187
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.15532633662223816,
      "learning_rate": 0.0001800583679231579,
      "loss": 1.1415,
      "step": 1188
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1169770359992981,
      "learning_rate": 0.0001800210084350081,
      "loss": 1.2081,
      "step": 1189
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13597778975963593,
      "learning_rate": 0.0001799836178676207,
      "loss": 1.1559,
      "step": 1190
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3288629949092865,
      "learning_rate": 0.00017994619623551777,
      "loss": 1.2292,
      "step": 1191
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13395251333713531,
      "learning_rate": 0.00017990874355323344,
      "loss": 1.2036,
      "step": 1192
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.10997981578111649,
      "learning_rate": 0.00017987125983531392,
      "loss": 1.1118,
      "step": 1193
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13059283792972565,
      "learning_rate": 0.00017983374509631742,
      "loss": 1.2046,
      "step": 1194
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12790393829345703,
      "learning_rate": 0.00017979619935081426,
      "loss": 1.1068,
      "step": 1195
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13489924371242523,
      "learning_rate": 0.0001797586226133867,
      "loss": 1.1125,
      "step": 1196
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1596228927373886,
      "learning_rate": 0.00017972101489862924,
      "loss": 1.106,
      "step": 1197
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1377343088388443,
      "learning_rate": 0.00017968337622114824,
      "loss": 1.0491,
      "step": 1198
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13254296779632568,
      "learning_rate": 0.00017964570659556207,
      "loss": 1.1246,
      "step": 1199
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1382158398628235,
      "learning_rate": 0.0001796080060365012,
      "loss": 1.1084,
      "step": 1200
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.10260984301567078,
      "learning_rate": 0.00017957027455860813,
      "loss": 1.0925,
      "step": 1201
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13929349184036255,
      "learning_rate": 0.0001795325121765373,
      "loss": 1.1923,
      "step": 1202
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.11720744520425797,
      "learning_rate": 0.00017949471890495522,
      "loss": 1.1985,
      "step": 1203
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12962299585342407,
      "learning_rate": 0.0001794568947585403,
      "loss": 1.0881,
      "step": 1204
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.251740038394928,
      "learning_rate": 0.00017941903975198306,
      "loss": 1.1389,
      "step": 1205
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12320899218320847,
      "learning_rate": 0.00017938115389998593,
      "loss": 1.1949,
      "step": 1206
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.15999193489551544,
      "learning_rate": 0.00017934323721726332,
      "loss": 1.1507,
      "step": 1207
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13963080942630768,
      "learning_rate": 0.00017930528971854163,
      "loss": 1.1409,
      "step": 1208
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1265689879655838,
      "learning_rate": 0.00017926731141855928,
      "loss": 1.1855,
      "step": 1209
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13507775962352753,
      "learning_rate": 0.00017922930233206657,
      "loss": 1.1093,
      "step": 1210
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.18033763766288757,
      "learning_rate": 0.00017919126247382575,
      "loss": 1.1368,
      "step": 1211
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12757407128810883,
      "learning_rate": 0.00017915319185861118,
      "loss": 1.1463,
      "step": 1212
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13044749200344086,
      "learning_rate": 0.0001791150905012089,
      "loss": 1.1436,
      "step": 1213
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.11629776656627655,
      "learning_rate": 0.00017907695841641713,
      "loss": 1.1837,
      "step": 1214
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1128249242901802,
      "learning_rate": 0.00017903879561904598,
      "loss": 1.2287,
      "step": 1215
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09792770445346832,
      "learning_rate": 0.0001790006021239173,
      "loss": 1.209,
      "step": 1216
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.14722852408885956,
      "learning_rate": 0.00017896237794586513,
      "loss": 1.1917,
      "step": 1217
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.13555045425891876,
      "learning_rate": 0.0001789241230997352,
      "loss": 1.1379,
      "step": 1218
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.14055520296096802,
      "learning_rate": 0.00017888583760038534,
      "loss": 1.1943,
      "step": 1219
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12172333151102066,
      "learning_rate": 0.00017884752146268512,
      "loss": 1.23,
      "step": 1220
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09420694410800934,
      "learning_rate": 0.00017880917470151615,
      "loss": 1.1916,
      "step": 1221
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1050577163696289,
      "learning_rate": 0.00017877079733177184,
      "loss": 1.1816,
      "step": 1222
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1305849850177765,
      "learning_rate": 0.0001787323893683575,
      "loss": 1.138,
      "step": 1223
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10311391949653625,
      "learning_rate": 0.0001786939508261904,
      "loss": 1.2474,
      "step": 1224
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10478417575359344,
      "learning_rate": 0.00017865548172019957,
      "loss": 1.0912,
      "step": 1225
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12189099937677383,
      "learning_rate": 0.00017861698206532596,
      "loss": 1.1258,
      "step": 1226
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.11864657700061798,
      "learning_rate": 0.00017857845187652246,
      "loss": 1.1529,
      "step": 1227
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.101483054459095,
      "learning_rate": 0.00017853989116875375,
      "loss": 1.1954,
      "step": 1228
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1290471851825714,
      "learning_rate": 0.00017850129995699624,
      "loss": 1.1267,
      "step": 1229
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.11972039192914963,
      "learning_rate": 0.00017846267825623845,
      "loss": 1.222,
      "step": 1230
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.13297469913959503,
      "learning_rate": 0.00017842402608148053,
      "loss": 1.1721,
      "step": 1231
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.14331290125846863,
      "learning_rate": 0.00017838534344773453,
      "loss": 1.1764,
      "step": 1232
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10414736717939377,
      "learning_rate": 0.00017834663037002443,
      "loss": 1.1998,
      "step": 1233
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.18757794797420502,
      "learning_rate": 0.00017830788686338584,
      "loss": 1.1756,
      "step": 1234
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.11975298821926117,
      "learning_rate": 0.00017826911294286637,
      "loss": 1.2032,
      "step": 1235
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12050428986549377,
      "learning_rate": 0.0001782303086235253,
      "loss": 1.1123,
      "step": 1236
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.13858585059642792,
      "learning_rate": 0.0001781914739204338,
      "loss": 1.148,
      "step": 1237
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.11384949088096619,
      "learning_rate": 0.00017815260884867484,
      "loss": 1.1218,
      "step": 1238
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12940974533557892,
      "learning_rate": 0.00017811371342334317,
      "loss": 1.1423,
      "step": 1239
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.09783143550157547,
      "learning_rate": 0.0001780747876595453,
      "loss": 1.1414,
      "step": 1240
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.131445974111557,
      "learning_rate": 0.00017803583157239956,
      "loss": 1.1553,
      "step": 1241
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12079741060733795,
      "learning_rate": 0.00017799684517703603,
      "loss": 1.2038,
      "step": 1242
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.11773856729269028,
      "learning_rate": 0.0001779578284885966,
      "loss": 1.2316,
      "step": 1243
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.15259861946105957,
      "learning_rate": 0.00017791878152223492,
      "loss": 1.1959,
      "step": 1244
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12402568757534027,
      "learning_rate": 0.00017787970429311632,
      "loss": 1.1299,
      "step": 1245
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10058469325304031,
      "learning_rate": 0.00017784059681641797,
      "loss": 1.1872,
      "step": 1246
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.11381770670413971,
      "learning_rate": 0.0001778014591073288,
      "loss": 1.1386,
      "step": 1247
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1401549130678177,
      "learning_rate": 0.00017776229118104942,
      "loss": 1.1682,
      "step": 1248
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10366252064704895,
      "learning_rate": 0.0001777230930527922,
      "loss": 1.2016,
      "step": 1249
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12752673029899597,
      "learning_rate": 0.00017768386473778123,
      "loss": 1.1808,
      "step": 1250
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.19915638864040375,
      "learning_rate": 0.00017764460625125235,
      "loss": 1.1324,
      "step": 1251
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.0838157907128334,
      "learning_rate": 0.0001776053176084531,
      "loss": 1.0576,
      "step": 1252
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.11519407480955124,
      "learning_rate": 0.00017756599882464273,
      "loss": 1.1684,
      "step": 1253
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1064494177699089,
      "learning_rate": 0.00017752664991509224,
      "loss": 1.15,
      "step": 1254
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1347271054983139,
      "learning_rate": 0.0001774872708950842,
      "loss": 1.1476,
      "step": 1255
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.13808561861515045,
      "learning_rate": 0.0001774478617799131,
      "loss": 1.173,
      "step": 1256
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.13670457899570465,
      "learning_rate": 0.00017740842258488488,
      "loss": 1.1566,
      "step": 1257
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1276087760925293,
      "learning_rate": 0.0001773689533253173,
      "loss": 1.1397,
      "step": 1258
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.13838480412960052,
      "learning_rate": 0.00017732945401653977,
      "loss": 1.1287,
      "step": 1259
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10592318326234818,
      "learning_rate": 0.0001772899246738934,
      "loss": 1.1481,
      "step": 1260
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.13724423944950104,
      "learning_rate": 0.00017725036531273088,
      "loss": 1.1497,
      "step": 1261
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12504073977470398,
      "learning_rate": 0.00017721077594841662,
      "loss": 1.0597,
      "step": 1262
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1512618213891983,
      "learning_rate": 0.00017717115659632667,
      "loss": 1.2772,
      "step": 1263
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.09324830025434494,
      "learning_rate": 0.00017713150727184876,
      "loss": 1.1006,
      "step": 1264
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12516917288303375,
      "learning_rate": 0.00017709182799038222,
      "loss": 1.1826,
      "step": 1265
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10511553287506104,
      "learning_rate": 0.000177052118767338,
      "loss": 1.1475,
      "step": 1266
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.08929230272769928,
      "learning_rate": 0.00017701237961813875,
      "loss": 1.0558,
      "step": 1267
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1099172905087471,
      "learning_rate": 0.00017697261055821863,
      "loss": 1.114,
      "step": 1268
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.11623919755220413,
      "learning_rate": 0.00017693281160302356,
      "loss": 1.1341,
      "step": 1269
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.12113003432750702,
      "learning_rate": 0.00017689298276801093,
      "loss": 1.1592,
      "step": 1270
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.16635805368423462,
      "learning_rate": 0.00017685312406864984,
      "loss": 1.1521,
      "step": 1271
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.13580547273159027,
      "learning_rate": 0.00017681323552042096,
      "loss": 1.1146,
      "step": 1272
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.08464888483285904,
      "learning_rate": 0.00017677331713881648,
      "loss": 1.0826,
      "step": 1273
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1189253181219101,
      "learning_rate": 0.0001767333689393403,
      "loss": 1.1193,
      "step": 1274
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11166702955961227,
      "learning_rate": 0.00017669339093750785,
      "loss": 1.1668,
      "step": 1275
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11939482390880585,
      "learning_rate": 0.00017665338314884607,
      "loss": 1.2184,
      "step": 1276
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.15603011846542358,
      "learning_rate": 0.00017661334558889357,
      "loss": 1.0973,
      "step": 1277
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.13399428129196167,
      "learning_rate": 0.00017657327827320045,
      "loss": 1.2038,
      "step": 1278
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1142265722155571,
      "learning_rate": 0.0001765331812173284,
      "loss": 1.191,
      "step": 1279
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1233569085597992,
      "learning_rate": 0.00017649305443685068,
      "loss": 1.1088,
      "step": 1280
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11068695038557053,
      "learning_rate": 0.000176452897947352,
      "loss": 1.0861,
      "step": 1281
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.13598217070102692,
      "learning_rate": 0.00017641271176442876,
      "loss": 1.2061,
      "step": 1282
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.12665072083473206,
      "learning_rate": 0.00017637249590368877,
      "loss": 1.1689,
      "step": 1283
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11580421030521393,
      "learning_rate": 0.00017633225038075142,
      "loss": 1.0643,
      "step": 1284
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.09630246460437775,
      "learning_rate": 0.00017629197521124755,
      "loss": 1.1766,
      "step": 1285
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.10641384869813919,
      "learning_rate": 0.00017625167041081965,
      "loss": 1.0834,
      "step": 1286
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.18533959984779358,
      "learning_rate": 0.00017621133599512163,
      "loss": 1.1648,
      "step": 1287
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11541745066642761,
      "learning_rate": 0.00017617097197981889,
      "loss": 1.113,
      "step": 1288
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.0872444361448288,
      "learning_rate": 0.0001761305783805883,
      "loss": 1.0393,
      "step": 1289
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.12431736290454865,
      "learning_rate": 0.00017609015521311835,
      "loss": 1.2045,
      "step": 1290
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1257573962211609,
      "learning_rate": 0.0001760497024931089,
      "loss": 1.1651,
      "step": 1291
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.10434801876544952,
      "learning_rate": 0.00017600922023627136,
      "loss": 1.1337,
      "step": 1292
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11347678303718567,
      "learning_rate": 0.0001759687084583285,
      "loss": 1.1829,
      "step": 1293
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.0990595892071724,
      "learning_rate": 0.0001759281671750147,
      "loss": 1.1206,
      "step": 1294
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.12281709909439087,
      "learning_rate": 0.00017588759640207563,
      "loss": 1.2036,
      "step": 1295
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.10643202811479568,
      "learning_rate": 0.00017584699615526858,
      "loss": 1.1018,
      "step": 1296
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11635127663612366,
      "learning_rate": 0.00017580636645036225,
      "loss": 1.2293,
      "step": 1297
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.10759002715349197,
      "learning_rate": 0.0001757657073031367,
      "loss": 1.1513,
      "step": 1298
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.12101371586322784,
      "learning_rate": 0.00017572501872938342,
      "loss": 1.1393,
      "step": 1299
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1356968730688095,
      "learning_rate": 0.0001756843007449055,
      "loss": 1.2129,
      "step": 1300
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.18132302165031433,
      "learning_rate": 0.00017564355336551727,
      "loss": 1.1345,
      "step": 1301
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.13688772916793823,
      "learning_rate": 0.00017560277660704453,
      "loss": 1.2253,
      "step": 1302
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1071641743183136,
      "learning_rate": 0.00017556197048532456,
      "loss": 1.1773,
      "step": 1303
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11106760799884796,
      "learning_rate": 0.00017552113501620594,
      "loss": 1.1403,
      "step": 1304
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1531020700931549,
      "learning_rate": 0.00017548027021554874,
      "loss": 1.0926,
      "step": 1305
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11437218636274338,
      "learning_rate": 0.0001754393760992243,
      "loss": 1.1142,
      "step": 1306
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11290910094976425,
      "learning_rate": 0.00017539845268311547,
      "loss": 1.1415,
      "step": 1307
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.10300295054912567,
      "learning_rate": 0.00017535749998311647,
      "loss": 1.1732,
      "step": 1308
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11862659454345703,
      "learning_rate": 0.0001753165180151328,
      "loss": 1.1818,
      "step": 1309
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.10293740034103394,
      "learning_rate": 0.0001752755067950814,
      "loss": 1.1961,
      "step": 1310
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1307038962841034,
      "learning_rate": 0.0001752344663388906,
      "loss": 1.1931,
      "step": 1311
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.13397730886936188,
      "learning_rate": 0.00017519339666249997,
      "loss": 1.2128,
      "step": 1312
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11857085675001144,
      "learning_rate": 0.00017515229778186054,
      "loss": 1.1454,
      "step": 1313
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11003360152244568,
      "learning_rate": 0.0001751111697129346,
      "loss": 1.1642,
      "step": 1314
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.12755653262138367,
      "learning_rate": 0.00017507001247169587,
      "loss": 1.147,
      "step": 1315
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.16620568931102753,
      "learning_rate": 0.00017502882607412933,
      "loss": 1.0465,
      "step": 1316
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.14502400159835815,
      "learning_rate": 0.00017498761053623128,
      "loss": 1.0732,
      "step": 1317
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1570545732975006,
      "learning_rate": 0.0001749463658740094,
      "loss": 1.0953,
      "step": 1318
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.13459515571594238,
      "learning_rate": 0.00017490509210348263,
      "loss": 1.122,
      "step": 1319
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1357787549495697,
      "learning_rate": 0.0001748637892406812,
      "loss": 1.0819,
      "step": 1320
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.12865345180034637,
      "learning_rate": 0.0001748224573016467,
      "loss": 1.2485,
      "step": 1321
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.15332293510437012,
      "learning_rate": 0.00017478109630243195,
      "loss": 1.1176,
      "step": 1322
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1307896077632904,
      "learning_rate": 0.0001747397062591011,
      "loss": 1.1079,
      "step": 1323
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1074582040309906,
      "learning_rate": 0.00017469828718772958,
      "loss": 1.134,
      "step": 1324
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.11506141722202301,
      "learning_rate": 0.00017465683910440405,
      "loss": 1.1036,
      "step": 1325
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.11054015904664993,
      "learning_rate": 0.00017461536202522247,
      "loss": 1.2941,
      "step": 1326
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14390164613723755,
      "learning_rate": 0.0001745738559662941,
      "loss": 1.194,
      "step": 1327
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14150145649909973,
      "learning_rate": 0.00017453232094373936,
      "loss": 1.1375,
      "step": 1328
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1586221307516098,
      "learning_rate": 0.00017449075697369005,
      "loss": 1.1613,
      "step": 1329
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.125178724527359,
      "learning_rate": 0.00017444916407228904,
      "loss": 1.1788,
      "step": 1330
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.12633655965328217,
      "learning_rate": 0.0001744075422556906,
      "loss": 1.1052,
      "step": 1331
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.15361426770687103,
      "learning_rate": 0.00017436589154006012,
      "loss": 1.183,
      "step": 1332
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14322367310523987,
      "learning_rate": 0.00017432421194157432,
      "loss": 1.163,
      "step": 1333
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.16189660131931305,
      "learning_rate": 0.000174282503476421,
      "loss": 1.2561,
      "step": 1334
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13292066752910614,
      "learning_rate": 0.00017424076616079934,
      "loss": 1.1661,
      "step": 1335
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13685783743858337,
      "learning_rate": 0.00017419900001091953,
      "loss": 1.1588,
      "step": 1336
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.12648828327655792,
      "learning_rate": 0.00017415720504300314,
      "loss": 1.1868,
      "step": 1337
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13209114968776703,
      "learning_rate": 0.00017411538127328283,
      "loss": 1.1256,
      "step": 1338
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14053253829479218,
      "learning_rate": 0.00017407352871800244,
      "loss": 1.2617,
      "step": 1339
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.11688601225614548,
      "learning_rate": 0.00017403164739341706,
      "loss": 1.1819,
      "step": 1340
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13821089267730713,
      "learning_rate": 0.0001739897373157929,
      "loss": 1.1211,
      "step": 1341
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13220161199569702,
      "learning_rate": 0.00017394779850140734,
      "loss": 1.1619,
      "step": 1342
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1265448033809662,
      "learning_rate": 0.00017390583096654896,
      "loss": 1.12,
      "step": 1343
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1400061547756195,
      "learning_rate": 0.00017386383472751743,
      "loss": 1.1731,
      "step": 1344
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1443544626235962,
      "learning_rate": 0.00017382180980062364,
      "loss": 1.1991,
      "step": 1345
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13031451404094696,
      "learning_rate": 0.00017377975620218953,
      "loss": 1.1497,
      "step": 1346
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.11938636749982834,
      "learning_rate": 0.00017373767394854834,
      "loss": 1.1468,
      "step": 1347
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.12663187086582184,
      "learning_rate": 0.00017369556305604422,
      "loss": 1.2059,
      "step": 1348
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1470104306936264,
      "learning_rate": 0.00017365342354103264,
      "loss": 1.136,
      "step": 1349
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.11949282884597778,
      "learning_rate": 0.00017361125541988002,
      "loss": 1.1417,
      "step": 1350
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14735174179077148,
      "learning_rate": 0.00017356905870896406,
      "loss": 1.1287,
      "step": 1351
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.11348456144332886,
      "learning_rate": 0.0001735268334246734,
      "loss": 1.1668,
      "step": 1352
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.12992914021015167,
      "learning_rate": 0.0001734845795834079,
      "loss": 1.145,
      "step": 1353
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.12392808496952057,
      "learning_rate": 0.00017344229720157845,
      "loss": 1.161,
      "step": 1354
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1333453208208084,
      "learning_rate": 0.00017339998629560706,
      "loss": 1.2244,
      "step": 1355
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14716681838035583,
      "learning_rate": 0.00017335764688192678,
      "loss": 1.147,
      "step": 1356
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.11014384031295776,
      "learning_rate": 0.0001733152789769817,
      "loss": 1.1971,
      "step": 1357
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.10939642786979675,
      "learning_rate": 0.00017327288259722713,
      "loss": 1.1168,
      "step": 1358
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.20381255447864532,
      "learning_rate": 0.00017323045775912926,
      "loss": 1.1388,
      "step": 1359
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.16126707196235657,
      "learning_rate": 0.00017318800447916542,
      "loss": 1.2164,
      "step": 1360
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.12315242737531662,
      "learning_rate": 0.00017314552277382402,
      "loss": 1.1304,
      "step": 1361
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.12961427867412567,
      "learning_rate": 0.00017310301265960446,
      "loss": 1.1835,
      "step": 1362
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14396464824676514,
      "learning_rate": 0.00017306047415301706,
      "loss": 1.2207,
      "step": 1363
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1236259937286377,
      "learning_rate": 0.00017301790727058345,
      "loss": 1.048,
      "step": 1364
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.11496347934007645,
      "learning_rate": 0.000172975312028836,
      "loss": 1.1608,
      "step": 1365
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.21699506044387817,
      "learning_rate": 0.00017293268844431828,
      "loss": 1.1969,
      "step": 1366
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.11854709684848785,
      "learning_rate": 0.00017289003653358473,
      "loss": 1.1025,
      "step": 1367
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13714483380317688,
      "learning_rate": 0.00017284735631320093,
      "loss": 1.1825,
      "step": 1368
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1396435797214508,
      "learning_rate": 0.00017280464779974335,
      "loss": 1.1605,
      "step": 1369
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13832566142082214,
      "learning_rate": 0.0001727619110097995,
      "loss": 1.1757,
      "step": 1370
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.12875424325466156,
      "learning_rate": 0.00017271914595996782,
      "loss": 1.1723,
      "step": 1371
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1311587691307068,
      "learning_rate": 0.00017267635266685782,
      "loss": 1.0924,
      "step": 1372
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.10631170868873596,
      "learning_rate": 0.0001726335311470899,
      "loss": 1.1437,
      "step": 1373
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13810501992702484,
      "learning_rate": 0.00017259068141729542,
      "loss": 1.227,
      "step": 1374
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14795319736003876,
      "learning_rate": 0.00017254780349411675,
      "loss": 1.2045,
      "step": 1375
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1411641240119934,
      "learning_rate": 0.00017250489739420718,
      "loss": 1.1123,
      "step": 1376
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.14148807525634766,
      "learning_rate": 0.00017246196313423093,
      "loss": 1.1449,
      "step": 1377
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.14077065885066986,
      "learning_rate": 0.00017241900073086317,
      "loss": 1.2025,
      "step": 1378
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12779203057289124,
      "learning_rate": 0.00017237601020079002,
      "loss": 1.2049,
      "step": 1379
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.15429385006427765,
      "learning_rate": 0.0001723329915607085,
      "loss": 1.1497,
      "step": 1380
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1308775544166565,
      "learning_rate": 0.00017228994482732651,
      "loss": 1.1621,
      "step": 1381
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.15433436632156372,
      "learning_rate": 0.000172246870017363,
      "loss": 1.1417,
      "step": 1382
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13014772534370422,
      "learning_rate": 0.00017220376714754764,
      "loss": 1.175,
      "step": 1383
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.11644397675991058,
      "learning_rate": 0.00017216063623462112,
      "loss": 1.1457,
      "step": 1384
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.09644661098718643,
      "learning_rate": 0.00017211747729533502,
      "loss": 1.116,
      "step": 1385
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12384694069623947,
      "learning_rate": 0.00017207429034645175,
      "loss": 1.1823,
      "step": 1386
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.14347052574157715,
      "learning_rate": 0.0001720310754047446,
      "loss": 1.121,
      "step": 1387
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12255574017763138,
      "learning_rate": 0.00017198783248699779,
      "loss": 1.1915,
      "step": 1388
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.11216936260461807,
      "learning_rate": 0.00017194456161000635,
      "loss": 1.1354,
      "step": 1389
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13430452346801758,
      "learning_rate": 0.0001719012627905762,
      "loss": 1.1659,
      "step": 1390
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.14243586361408234,
      "learning_rate": 0.00017185793604552408,
      "loss": 1.0961,
      "step": 1391
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13742303848266602,
      "learning_rate": 0.00017181458139167764,
      "loss": 1.163,
      "step": 1392
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.15447835624217987,
      "learning_rate": 0.00017177119884587535,
      "loss": 1.0951,
      "step": 1393
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1354411393404007,
      "learning_rate": 0.0001717277884249664,
      "loss": 1.1877,
      "step": 1394
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12019835412502289,
      "learning_rate": 0.00017168435014581094,
      "loss": 1.1542,
      "step": 1395
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1273394227027893,
      "learning_rate": 0.0001716408840252799,
      "loss": 1.1612,
      "step": 1396
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.15226763486862183,
      "learning_rate": 0.00017159739008025505,
      "loss": 1.2028,
      "step": 1397
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.10990821570158005,
      "learning_rate": 0.0001715538683276289,
      "loss": 1.1269,
      "step": 1398
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1353336125612259,
      "learning_rate": 0.0001715103187843048,
      "loss": 1.0541,
      "step": 1399
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.16755186021327972,
      "learning_rate": 0.0001714667414671969,
      "loss": 1.1534,
      "step": 1400
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.14377646148204803,
      "learning_rate": 0.0001714231363932301,
      "loss": 1.2214,
      "step": 1401
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.17470446228981018,
      "learning_rate": 0.00017137950357934016,
      "loss": 1.1519,
      "step": 1402
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13566303253173828,
      "learning_rate": 0.00017133584304247353,
      "loss": 1.2139,
      "step": 1403
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.11175883561372757,
      "learning_rate": 0.00017129215479958745,
      "loss": 1.1338,
      "step": 1404
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13073724508285522,
      "learning_rate": 0.00017124843886765,
      "loss": 1.2254,
      "step": 1405
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12734417617321014,
      "learning_rate": 0.00017120469526363982,
      "loss": 1.0939,
      "step": 1406
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.15097704529762268,
      "learning_rate": 0.00017116092400454652,
      "loss": 1.1469,
      "step": 1407
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.11862929910421371,
      "learning_rate": 0.00017111712510737035,
      "loss": 1.1524,
      "step": 1408
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.10930761694908142,
      "learning_rate": 0.00017107329858912225,
      "loss": 1.2018,
      "step": 1409
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.131627157330513,
      "learning_rate": 0.00017102944446682395,
      "loss": 1.1704,
      "step": 1410
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1377275437116623,
      "learning_rate": 0.00017098556275750787,
      "loss": 1.1244,
      "step": 1411
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.11810161918401718,
      "learning_rate": 0.00017094165347821723,
      "loss": 1.1264,
      "step": 1412
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13606129586696625,
      "learning_rate": 0.00017089771664600582,
      "loss": 1.1409,
      "step": 1413
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.20456425845623016,
      "learning_rate": 0.00017085375227793818,
      "loss": 1.0319,
      "step": 1414
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.11501086503267288,
      "learning_rate": 0.00017080976039108965,
      "loss": 1.0823,
      "step": 1415
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.14591048657894135,
      "learning_rate": 0.00017076574100254613,
      "loss": 1.1498,
      "step": 1416
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.14271678030490875,
      "learning_rate": 0.00017072169412940422,
      "loss": 1.1365,
      "step": 1417
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12837226688861847,
      "learning_rate": 0.00017067761978877121,
      "loss": 1.1857,
      "step": 1418
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13796739280223846,
      "learning_rate": 0.00017063351799776513,
      "loss": 1.1453,
      "step": 1419
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13679030537605286,
      "learning_rate": 0.00017058938877351456,
      "loss": 1.2044,
      "step": 1420
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1605735868215561,
      "learning_rate": 0.00017054523213315878,
      "loss": 1.0948,
      "step": 1421
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.10744545608758926,
      "learning_rate": 0.00017050104809384774,
      "loss": 1.1461,
      "step": 1422
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.11329437047243118,
      "learning_rate": 0.000170456836672742,
      "loss": 1.1794,
      "step": 1423
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.1548263281583786,
      "learning_rate": 0.00017041259788701281,
      "loss": 1.2471,
      "step": 1424
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.13597482442855835,
      "learning_rate": 0.0001703683317538419,
      "loss": 1.1251,
      "step": 1425
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12478166073560715,
      "learning_rate": 0.00017032403829042182,
      "loss": 1.1774,
      "step": 1426
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.14619958400726318,
      "learning_rate": 0.00017027971751395563,
      "loss": 1.2238,
      "step": 1427
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.0998232364654541,
      "learning_rate": 0.00017023536944165698,
      "loss": 1.1304,
      "step": 1428
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.12032195180654526,
      "learning_rate": 0.00017019099409075014,
      "loss": 1.1247,
      "step": 1429
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1156492829322815,
      "learning_rate": 0.00017014659147847002,
      "loss": 1.1246,
      "step": 1430
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.13250748813152313,
      "learning_rate": 0.00017010216162206208,
      "loss": 1.0753,
      "step": 1431
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.13088637590408325,
      "learning_rate": 0.00017005770453878234,
      "loss": 1.1855,
      "step": 1432
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.15802554786205292,
      "learning_rate": 0.00017001322024589742,
      "loss": 1.0603,
      "step": 1433
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.14481595158576965,
      "learning_rate": 0.00016996870876068452,
      "loss": 1.1661,
      "step": 1434
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.13875627517700195,
      "learning_rate": 0.00016992417010043142,
      "loss": 1.1088,
      "step": 1435
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12186279892921448,
      "learning_rate": 0.00016987960428243638,
      "loss": 1.152,
      "step": 1436
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1580241173505783,
      "learning_rate": 0.00016983501132400824,
      "loss": 1.1864,
      "step": 1437
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12886583805084229,
      "learning_rate": 0.00016979039124246642,
      "loss": 1.1834,
      "step": 1438
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.14764368534088135,
      "learning_rate": 0.00016974574405514083,
      "loss": 1.1289,
      "step": 1439
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1422930508852005,
      "learning_rate": 0.00016970106977937192,
      "loss": 1.1426,
      "step": 1440
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.10564149171113968,
      "learning_rate": 0.0001696563684325107,
      "loss": 1.1664,
      "step": 1441
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1388014405965805,
      "learning_rate": 0.00016961164003191862,
      "loss": 1.2091,
      "step": 1442
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.14382459223270416,
      "learning_rate": 0.00016956688459496765,
      "loss": 1.1485,
      "step": 1443
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1639106571674347,
      "learning_rate": 0.00016952210213904038,
      "loss": 1.1952,
      "step": 1444
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.13041575253009796,
      "learning_rate": 0.0001694772926815297,
      "loss": 1.1265,
      "step": 1445
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12336768954992294,
      "learning_rate": 0.00016943245623983916,
      "loss": 1.1179,
      "step": 1446
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12504109740257263,
      "learning_rate": 0.00016938759283138268,
      "loss": 1.1843,
      "step": 1447
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12632761895656586,
      "learning_rate": 0.00016934270247358468,
      "loss": 1.1152,
      "step": 1448
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.14493171870708466,
      "learning_rate": 0.00016929778518388007,
      "loss": 1.1909,
      "step": 1449
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.10036037117242813,
      "learning_rate": 0.00016925284097971425,
      "loss": 1.1255,
      "step": 1450
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12682358920574188,
      "learning_rate": 0.00016920786987854294,
      "loss": 1.0986,
      "step": 1451
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.16553513705730438,
      "learning_rate": 0.00016916287189783252,
      "loss": 1.1409,
      "step": 1452
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.14230042695999146,
      "learning_rate": 0.00016911784705505957,
      "loss": 1.0965,
      "step": 1453
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1273324340581894,
      "learning_rate": 0.0001690727953677113,
      "loss": 1.2159,
      "step": 1454
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11146815121173859,
      "learning_rate": 0.00016902771685328523,
      "loss": 1.0516,
      "step": 1455
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.163333460688591,
      "learning_rate": 0.00016898261152928931,
      "loss": 1.2301,
      "step": 1456
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.09529362618923187,
      "learning_rate": 0.00016893747941324196,
      "loss": 0.9997,
      "step": 1457
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11501886695623398,
      "learning_rate": 0.000168892320522672,
      "loss": 1.2032,
      "step": 1458
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.16507099568843842,
      "learning_rate": 0.00016884713487511858,
      "loss": 1.1276,
      "step": 1459
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11970765143632889,
      "learning_rate": 0.0001688019224881313,
      "loss": 1.1203,
      "step": 1460
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12286444753408432,
      "learning_rate": 0.00016875668337927013,
      "loss": 1.1989,
      "step": 1461
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11843866854906082,
      "learning_rate": 0.00016871141756610544,
      "loss": 1.1693,
      "step": 1462
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.09849977493286133,
      "learning_rate": 0.0001686661250662179,
      "loss": 1.2516,
      "step": 1463
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.14754219353199005,
      "learning_rate": 0.0001686208058971986,
      "loss": 1.1777,
      "step": 1464
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11357257515192032,
      "learning_rate": 0.00016857546007664905,
      "loss": 1.1768,
      "step": 1465
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11370682716369629,
      "learning_rate": 0.000168530087622181,
      "loss": 1.2164,
      "step": 1466
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1377449631690979,
      "learning_rate": 0.0001684846885514166,
      "loss": 1.0966,
      "step": 1467
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.12039929628372192,
      "learning_rate": 0.00016843926288198828,
      "loss": 1.0742,
      "step": 1468
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.15042871236801147,
      "learning_rate": 0.0001683938106315389,
      "loss": 1.096,
      "step": 1469
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1563349813222885,
      "learning_rate": 0.0001683483318177216,
      "loss": 1.2004,
      "step": 1470
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1320793479681015,
      "learning_rate": 0.00016830282645819976,
      "loss": 1.1346,
      "step": 1471
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1512555032968521,
      "learning_rate": 0.0001682572945706472,
      "loss": 1.1501,
      "step": 1472
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11943696439266205,
      "learning_rate": 0.00016821173617274793,
      "loss": 1.2042,
      "step": 1473
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1985544115304947,
      "learning_rate": 0.00016816615128219635,
      "loss": 1.0933,
      "step": 1474
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11811183393001556,
      "learning_rate": 0.0001681205399166971,
      "loss": 1.1584,
      "step": 1475
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.15123914182186127,
      "learning_rate": 0.00016807490209396506,
      "loss": 1.1299,
      "step": 1476
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1436377763748169,
      "learning_rate": 0.00016802923783172552,
      "loss": 1.1316,
      "step": 1477
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.14886842668056488,
      "learning_rate": 0.0001679835471477139,
      "loss": 1.1835,
      "step": 1478
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.14267393946647644,
      "learning_rate": 0.00016793783005967592,
      "loss": 1.1595,
      "step": 1479
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1237650141119957,
      "learning_rate": 0.0001678920865853676,
      "loss": 1.1767,
      "step": 1480
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.11216934025287628,
      "learning_rate": 0.00016784631674255518,
      "loss": 1.1394,
      "step": 1481
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1582978516817093,
      "learning_rate": 0.00016780052054901512,
      "loss": 1.2258,
      "step": 1482
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.16898638010025024,
      "learning_rate": 0.00016775469802253414,
      "loss": 1.196,
      "step": 1483
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1693446934223175,
      "learning_rate": 0.0001677088491809092,
      "loss": 1.0858,
      "step": 1484
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.16614001989364624,
      "learning_rate": 0.00016766297404194745,
      "loss": 1.1752,
      "step": 1485
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.11463948339223862,
      "learning_rate": 0.00016761707262346625,
      "loss": 1.1933,
      "step": 1486
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.11729386448860168,
      "learning_rate": 0.0001675711449432932,
      "loss": 1.1042,
      "step": 1487
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.15680693089962006,
      "learning_rate": 0.00016752519101926607,
      "loss": 1.2198,
      "step": 1488
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12007825821638107,
      "learning_rate": 0.00016747921086923283,
      "loss": 1.1638,
      "step": 1489
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.146340012550354,
      "learning_rate": 0.00016743320451105168,
      "loss": 1.1019,
      "step": 1490
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.10307899117469788,
      "learning_rate": 0.00016738717196259094,
      "loss": 1.1389,
      "step": 1491
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.13613498210906982,
      "learning_rate": 0.0001673411132417291,
      "loss": 1.141,
      "step": 1492
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1158485934138298,
      "learning_rate": 0.0001672950283663548,
      "loss": 1.1922,
      "step": 1493
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.15027955174446106,
      "learning_rate": 0.00016724891735436697,
      "loss": 1.0808,
      "step": 1494
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.11840091645717621,
      "learning_rate": 0.00016720278022367452,
      "loss": 1.2019,
      "step": 1495
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12492614984512329,
      "learning_rate": 0.00016715661699219663,
      "loss": 1.1315,
      "step": 1496
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.11582069098949432,
      "learning_rate": 0.00016711042767786257,
      "loss": 1.1509,
      "step": 1497
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.09216103702783585,
      "learning_rate": 0.00016706421229861166,
      "loss": 1.118,
      "step": 1498
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.10561263561248779,
      "learning_rate": 0.00016701797087239354,
      "loss": 1.1821,
      "step": 1499
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.13040509819984436,
      "learning_rate": 0.0001669717034171677,
      "loss": 1.1017,
      "step": 1500
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.11197645962238312,
      "learning_rate": 0.000166925409950904,
      "loss": 1.274,
      "step": 1501
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3740244209766388,
      "learning_rate": 0.00016687909049158228,
      "loss": 1.1305,
      "step": 1502
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.13568273186683655,
      "learning_rate": 0.00016683274505719246,
      "loss": 1.1465,
      "step": 1503
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.11974001675844193,
      "learning_rate": 0.00016678637366573454,
      "loss": 1.0481,
      "step": 1504
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.16084937751293182,
      "learning_rate": 0.0001667399763352187,
      "loss": 1.1688,
      "step": 1505
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.153037890791893,
      "learning_rate": 0.0001666935530836651,
      "loss": 1.2294,
      "step": 1506
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.11359276622533798,
      "learning_rate": 0.00016664710392910395,
      "loss": 1.1415,
      "step": 1507
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.09300748258829117,
      "learning_rate": 0.00016660062888957563,
      "loss": 1.139,
      "step": 1508
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.10428538173437119,
      "learning_rate": 0.00016655412798313051,
      "loss": 1.1189,
      "step": 1509
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12125095725059509,
      "learning_rate": 0.00016650760122782895,
      "loss": 1.1821,
      "step": 1510
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.14301297068595886,
      "learning_rate": 0.00016646104864174145,
      "loss": 1.1053,
      "step": 1511
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1356104016304016,
      "learning_rate": 0.0001664144702429485,
      "loss": 1.1287,
      "step": 1512
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1283227652311325,
      "learning_rate": 0.0001663678660495406,
      "loss": 1.1689,
      "step": 1513
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.15683116018772125,
      "learning_rate": 0.0001663212360796183,
      "loss": 1.217,
      "step": 1514
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.11689150333404541,
      "learning_rate": 0.0001662745803512921,
      "loss": 1.1805,
      "step": 1515
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12443431466817856,
      "learning_rate": 0.00016622789888268258,
      "loss": 1.1805,
      "step": 1516
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1745922565460205,
      "learning_rate": 0.00016618119169192026,
      "loss": 1.1957,
      "step": 1517
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12343040108680725,
      "learning_rate": 0.00016613445879714572,
      "loss": 1.1405,
      "step": 1518
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.14226172864437103,
      "learning_rate": 0.00016608770021650943,
      "loss": 1.1513,
      "step": 1519
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.1273631602525711,
      "learning_rate": 0.00016604091596817192,
      "loss": 1.2065,
      "step": 1520
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.17296408116817474,
      "learning_rate": 0.00016599410607030365,
      "loss": 1.1258,
      "step": 1521
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.15095782279968262,
      "learning_rate": 0.00016594727054108498,
      "loss": 1.0596,
      "step": 1522
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.15684297680854797,
      "learning_rate": 0.0001659004093987064,
      "loss": 1.0973,
      "step": 1523
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12869741022586823,
      "learning_rate": 0.00016585352266136814,
      "loss": 1.1491,
      "step": 1524
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12245168536901474,
      "learning_rate": 0.00016580661034728053,
      "loss": 1.2512,
      "step": 1525
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.11353924125432968,
      "learning_rate": 0.00016575967247466376,
      "loss": 1.1523,
      "step": 1526
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.14712193608283997,
      "learning_rate": 0.0001657127090617479,
      "loss": 1.112,
      "step": 1527
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12008864432573318,
      "learning_rate": 0.0001656657201267731,
      "loss": 1.106,
      "step": 1528
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.10149414092302322,
      "learning_rate": 0.00016561870568798924,
      "loss": 1.1741,
      "step": 1529
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.131751149892807,
      "learning_rate": 0.00016557166576365622,
      "loss": 1.2614,
      "step": 1530
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.21924680471420288,
      "learning_rate": 0.00016552460037204384,
      "loss": 1.114,
      "step": 1531
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.13210196793079376,
      "learning_rate": 0.00016547750953143167,
      "loss": 1.1781,
      "step": 1532
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.13061179220676422,
      "learning_rate": 0.00016543039326010928,
      "loss": 1.1208,
      "step": 1533
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12237674742937088,
      "learning_rate": 0.00016538325157637614,
      "loss": 1.1044,
      "step": 1534
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.13162200152873993,
      "learning_rate": 0.00016533608449854147,
      "loss": 1.0911,
      "step": 1535
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.17683564126491547,
      "learning_rate": 0.00016528889204492448,
      "loss": 1.2487,
      "step": 1536
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12894952297210693,
      "learning_rate": 0.00016524167423385413,
      "loss": 1.0894,
      "step": 1537
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1351403295993805,
      "learning_rate": 0.00016519443108366927,
      "loss": 1.1496,
      "step": 1538
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1572115272283554,
      "learning_rate": 0.00016514716261271866,
      "loss": 1.1826,
      "step": 1539
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.128783717751503,
      "learning_rate": 0.00016509986883936074,
      "loss": 1.086,
      "step": 1540
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.15486739575862885,
      "learning_rate": 0.00016505254978196388,
      "loss": 1.2348,
      "step": 1541
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1370764523744583,
      "learning_rate": 0.00016500520545890634,
      "loss": 1.1945,
      "step": 1542
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1305244117975235,
      "learning_rate": 0.00016495783588857605,
      "loss": 1.2154,
      "step": 1543
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1454448103904724,
      "learning_rate": 0.0001649104410893708,
      "loss": 1.039,
      "step": 1544
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.14552029967308044,
      "learning_rate": 0.0001648630210796982,
      "loss": 1.2616,
      "step": 1545
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.149075448513031,
      "learning_rate": 0.0001648155758779756,
      "loss": 1.1149,
      "step": 1546
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12380526959896088,
      "learning_rate": 0.0001647681055026302,
      "loss": 1.0713,
      "step": 1547
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11009863018989563,
      "learning_rate": 0.000164720609972099,
      "loss": 1.1594,
      "step": 1548
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.09933313727378845,
      "learning_rate": 0.00016467308930482864,
      "loss": 1.1323,
      "step": 1549
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11832248419523239,
      "learning_rate": 0.00016462554351927557,
      "loss": 1.1869,
      "step": 1550
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.15750133991241455,
      "learning_rate": 0.00016457797263390612,
      "loss": 1.1179,
      "step": 1551
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.16294234991073608,
      "learning_rate": 0.00016453037666719624,
      "loss": 1.1926,
      "step": 1552
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11820466816425323,
      "learning_rate": 0.00016448275563763162,
      "loss": 1.1401,
      "step": 1553
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.10791966319084167,
      "learning_rate": 0.0001644351095637078,
      "loss": 1.1471,
      "step": 1554
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.13995571434497833,
      "learning_rate": 0.00016438743846392985,
      "loss": 1.1897,
      "step": 1555
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12756232917308807,
      "learning_rate": 0.00016433974235681274,
      "loss": 1.1531,
      "step": 1556
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.13782839477062225,
      "learning_rate": 0.0001642920212608811,
      "loss": 1.1916,
      "step": 1557
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.13091681897640228,
      "learning_rate": 0.00016424427519466924,
      "loss": 1.166,
      "step": 1558
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12873682379722595,
      "learning_rate": 0.00016419650417672118,
      "loss": 1.1619,
      "step": 1559
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1430850625038147,
      "learning_rate": 0.00016414870822559064,
      "loss": 1.1407,
      "step": 1560
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.10752367228269577,
      "learning_rate": 0.00016410088735984102,
      "loss": 1.105,
      "step": 1561
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12413575500249863,
      "learning_rate": 0.00016405304159804534,
      "loss": 1.1301,
      "step": 1562
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12152840942144394,
      "learning_rate": 0.00016400517095878643,
      "loss": 1.1912,
      "step": 1563
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.10437405854463577,
      "learning_rate": 0.00016395727546065665,
      "loss": 1.1178,
      "step": 1564
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11455471068620682,
      "learning_rate": 0.00016390935512225805,
      "loss": 1.1263,
      "step": 1565
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.10488203167915344,
      "learning_rate": 0.00016386140996220232,
      "loss": 1.1556,
      "step": 1566
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1322120726108551,
      "learning_rate": 0.00016381343999911086,
      "loss": 1.1705,
      "step": 1567
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.19445818662643433,
      "learning_rate": 0.00016376544525161465,
      "loss": 1.1066,
      "step": 1568
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11544685810804367,
      "learning_rate": 0.00016371742573835426,
      "loss": 1.1142,
      "step": 1569
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1091468408703804,
      "learning_rate": 0.0001636693814779799,
      "loss": 1.1267,
      "step": 1570
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1646507978439331,
      "learning_rate": 0.00016362131248915144,
      "loss": 1.1444,
      "step": 1571
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11092719435691833,
      "learning_rate": 0.00016357321879053834,
      "loss": 1.1486,
      "step": 1572
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11516217887401581,
      "learning_rate": 0.00016352510040081963,
      "loss": 1.2397,
      "step": 1573
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12745524942874908,
      "learning_rate": 0.00016347695733868388,
      "loss": 0.9921,
      "step": 1574
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1421622633934021,
      "learning_rate": 0.00016342878962282938,
      "loss": 1.1426,
      "step": 1575
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.09525369852781296,
      "learning_rate": 0.00016338059727196387,
      "loss": 1.0992,
      "step": 1576
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12791882455348969,
      "learning_rate": 0.0001633323803048047,
      "loss": 1.1978,
      "step": 1577
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12086856365203857,
      "learning_rate": 0.0001632841387400788,
      "loss": 1.1955,
      "step": 1578
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1353377103805542,
      "learning_rate": 0.00016323587259652267,
      "loss": 1.1616,
      "step": 1579
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.13189376890659332,
      "learning_rate": 0.00016318758189288226,
      "loss": 1.2329,
      "step": 1580
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.12470117956399918,
      "learning_rate": 0.00016313926664791314,
      "loss": 1.1566,
      "step": 1581
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11324245482683182,
      "learning_rate": 0.00016309092688038046,
      "loss": 1.2449,
      "step": 1582
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.14797192811965942,
      "learning_rate": 0.00016304256260905872,
      "loss": 1.1353,
      "step": 1583
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.10740874707698822,
      "learning_rate": 0.00016299417385273214,
      "loss": 1.1356,
      "step": 1584
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.11877727508544922,
      "learning_rate": 0.00016294576063019428,
      "loss": 1.1651,
      "step": 1585
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11185145378112793,
      "learning_rate": 0.00016289732296024834,
      "loss": 1.1007,
      "step": 1586
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.12083572894334793,
      "learning_rate": 0.00016284886086170698,
      "loss": 1.1422,
      "step": 1587
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11183588951826096,
      "learning_rate": 0.00016280037435339217,
      "loss": 1.1228,
      "step": 1588
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.14046955108642578,
      "learning_rate": 0.00016275186345413568,
      "loss": 1.1535,
      "step": 1589
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.16720926761627197,
      "learning_rate": 0.00016270332818277847,
      "loss": 1.2225,
      "step": 1590
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1849929839372635,
      "learning_rate": 0.00016265476855817116,
      "loss": 1.2053,
      "step": 1591
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11735444515943527,
      "learning_rate": 0.00016260618459917368,
      "loss": 1.1428,
      "step": 1592
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.10270600765943527,
      "learning_rate": 0.00016255757632465553,
      "loss": 1.1006,
      "step": 1593
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1113390102982521,
      "learning_rate": 0.00016250894375349558,
      "loss": 1.0987,
      "step": 1594
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13255318999290466,
      "learning_rate": 0.00016246028690458216,
      "loss": 1.0746,
      "step": 1595
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.14710868895053864,
      "learning_rate": 0.00016241160579681308,
      "loss": 1.0933,
      "step": 1596
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.16927458345890045,
      "learning_rate": 0.0001623629004490954,
      "loss": 1.0537,
      "step": 1597
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.18780843913555145,
      "learning_rate": 0.00016231417088034586,
      "loss": 1.181,
      "step": 1598
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.14640431106090546,
      "learning_rate": 0.0001622654171094904,
      "loss": 1.221,
      "step": 1599
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13740523159503937,
      "learning_rate": 0.00016221663915546436,
      "loss": 1.0624,
      "step": 1600
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.14213408529758453,
      "learning_rate": 0.00016216783703721266,
      "loss": 1.1654,
      "step": 1601
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13336369395256042,
      "learning_rate": 0.00016211901077368935,
      "loss": 1.1695,
      "step": 1602
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1228618174791336,
      "learning_rate": 0.0001620701603838581,
      "loss": 1.1804,
      "step": 1603
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13464073836803436,
      "learning_rate": 0.00016202128588669177,
      "loss": 1.1448,
      "step": 1604
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.12865209579467773,
      "learning_rate": 0.00016197238730117271,
      "loss": 1.2118,
      "step": 1605
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11904550343751907,
      "learning_rate": 0.00016192346464629246,
      "loss": 1.2165,
      "step": 1606
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11649783700704575,
      "learning_rate": 0.00016187451794105214,
      "loss": 1.1579,
      "step": 1607
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11373747140169144,
      "learning_rate": 0.00016182554720446202,
      "loss": 1.1418,
      "step": 1608
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13193392753601074,
      "learning_rate": 0.00016177655245554177,
      "loss": 1.1841,
      "step": 1609
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.14541228115558624,
      "learning_rate": 0.00016172753371332038,
      "loss": 1.2344,
      "step": 1610
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13203716278076172,
      "learning_rate": 0.00016167849099683624,
      "loss": 1.11,
      "step": 1611
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1387616991996765,
      "learning_rate": 0.00016162942432513687,
      "loss": 1.1133,
      "step": 1612
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11879047751426697,
      "learning_rate": 0.00016158033371727924,
      "loss": 1.1414,
      "step": 1613
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.15034838020801544,
      "learning_rate": 0.00016153121919232962,
      "loss": 1.1435,
      "step": 1614
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.09441407769918442,
      "learning_rate": 0.00016148208076936348,
      "loss": 1.2066,
      "step": 1615
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1158721074461937,
      "learning_rate": 0.0001614329184674656,
      "loss": 1.1094,
      "step": 1616
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.09111738950014114,
      "learning_rate": 0.00016138373230573013,
      "loss": 1.0794,
      "step": 1617
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11628139019012451,
      "learning_rate": 0.00016133452230326033,
      "loss": 1.0698,
      "step": 1618
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.15395097434520721,
      "learning_rate": 0.00016128528847916883,
      "loss": 1.1739,
      "step": 1619
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.12554185092449188,
      "learning_rate": 0.00016123603085257745,
      "loss": 1.141,
      "step": 1620
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.14405061304569244,
      "learning_rate": 0.00016118674944261732,
      "loss": 1.149,
      "step": 1621
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.16187065839767456,
      "learning_rate": 0.0001611374442684288,
      "loss": 1.1554,
      "step": 1622
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.0982840359210968,
      "learning_rate": 0.00016108811534916136,
      "loss": 1.1467,
      "step": 1623
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.10352392494678497,
      "learning_rate": 0.00016103876270397386,
      "loss": 1.1523,
      "step": 1624
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11771900206804276,
      "learning_rate": 0.0001609893863520343,
      "loss": 1.1372,
      "step": 1625
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.1788703203201294,
      "learning_rate": 0.0001609399863125198,
      "loss": 1.2087,
      "step": 1626
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.10901201516389847,
      "learning_rate": 0.00016089056260461688,
      "loss": 1.1722,
      "step": 1627
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.15400920808315277,
      "learning_rate": 0.00016084111524752105,
      "loss": 1.0516,
      "step": 1628
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.17698659002780914,
      "learning_rate": 0.00016079164426043718,
      "loss": 1.0765,
      "step": 1629
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.12125472724437714,
      "learning_rate": 0.0001607421496625791,
      "loss": 1.2078,
      "step": 1630
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.12957288324832916,
      "learning_rate": 0.00016069263147317013,
      "loss": 1.1115,
      "step": 1631
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.10030249506235123,
      "learning_rate": 0.00016064308971144238,
      "loss": 1.1943,
      "step": 1632
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.10950974375009537,
      "learning_rate": 0.00016059352439663739,
      "loss": 1.1046,
      "step": 1633
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13490870594978333,
      "learning_rate": 0.00016054393554800576,
      "loss": 1.115,
      "step": 1634
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.13052257895469666,
      "learning_rate": 0.0001604943231848072,
      "loss": 1.1467,
      "step": 1635
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.14307262003421783,
      "learning_rate": 0.00016044468732631057,
      "loss": 1.2375,
      "step": 1636
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.11784642934799194,
      "learning_rate": 0.00016039502799179394,
      "loss": 1.1223,
      "step": 1637
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.13222099840641022,
      "learning_rate": 0.00016034534520054433,
      "loss": 1.1154,
      "step": 1638
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.12012042850255966,
      "learning_rate": 0.000160295638971858,
      "loss": 1.151,
      "step": 1639
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.14703696966171265,
      "learning_rate": 0.0001602459093250403,
      "loss": 1.136,
      "step": 1640
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.15325959026813507,
      "learning_rate": 0.0001601961562794056,
      "loss": 1.2507,
      "step": 1641
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.12960705161094666,
      "learning_rate": 0.0001601463798542775,
      "loss": 1.134,
      "step": 1642
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.10418795049190521,
      "learning_rate": 0.00016009658006898848,
      "loss": 1.156,
      "step": 1643
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.11279977858066559,
      "learning_rate": 0.00016004675694288026,
      "loss": 1.1587,
      "step": 1644
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.145701065659523,
      "learning_rate": 0.0001599969104953036,
      "loss": 1.104,
      "step": 1645
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.305241346359253,
      "learning_rate": 0.0001599470407456182,
      "loss": 1.1101,
      "step": 1646
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1747491955757141,
      "learning_rate": 0.00015989714771319299,
      "loss": 1.122,
      "step": 1647
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.11683280766010284,
      "learning_rate": 0.00015984723141740576,
      "loss": 1.1761,
      "step": 1648
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1230531632900238,
      "learning_rate": 0.0001597972918776435,
      "loss": 1.117,
      "step": 1649
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1510223150253296,
      "learning_rate": 0.00015974732911330208,
      "loss": 1.144,
      "step": 1650
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.14948098361492157,
      "learning_rate": 0.00015969734314378654,
      "loss": 1.0494,
      "step": 1651
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09652502834796906,
      "learning_rate": 0.00015964733398851077,
      "loss": 1.0841,
      "step": 1652
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09940962493419647,
      "learning_rate": 0.00015959730166689783,
      "loss": 1.0988,
      "step": 1653
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1498783528804779,
      "learning_rate": 0.00015954724619837967,
      "loss": 1.1482,
      "step": 1654
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.12323389947414398,
      "learning_rate": 0.00015949716760239722,
      "loss": 1.2122,
      "step": 1655
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1720038503408432,
      "learning_rate": 0.00015944706589840046,
      "loss": 1.1327,
      "step": 1656
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.17082177102565765,
      "learning_rate": 0.00015939694110584832,
      "loss": 1.0919,
      "step": 1657
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.16806374490261078,
      "learning_rate": 0.00015934679324420872,
      "loss": 1.1991,
      "step": 1658
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.14044439792633057,
      "learning_rate": 0.00015929662233295843,
      "loss": 1.2509,
      "step": 1659
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.11187873780727386,
      "learning_rate": 0.00015924642839158332,
      "loss": 1.1453,
      "step": 1660
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.11407248675823212,
      "learning_rate": 0.0001591962114395781,
      "loss": 1.1585,
      "step": 1661
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.10827763378620148,
      "learning_rate": 0.00015914597149644652,
      "loss": 1.1647,
      "step": 1662
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.16270461678504944,
      "learning_rate": 0.00015909570858170112,
      "loss": 1.2163,
      "step": 1663
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.12677918374538422,
      "learning_rate": 0.00015904542271486346,
      "loss": 1.2212,
      "step": 1664
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1433684378862381,
      "learning_rate": 0.00015899511391546402,
      "loss": 1.1511,
      "step": 1665
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1129528060555458,
      "learning_rate": 0.00015894478220304214,
      "loss": 1.1361,
      "step": 1666
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.14624518156051636,
      "learning_rate": 0.00015889442759714603,
      "loss": 1.2252,
      "step": 1667
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1491534262895584,
      "learning_rate": 0.00015884405011733292,
      "loss": 1.0613,
      "step": 1668
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.20824897289276123,
      "learning_rate": 0.0001587936497831688,
      "loss": 1.1485,
      "step": 1669
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1260470151901245,
      "learning_rate": 0.00015874322661422856,
      "loss": 1.0816,
      "step": 1670
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.14912791550159454,
      "learning_rate": 0.000158692780630096,
      "loss": 1.1997,
      "step": 1671
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.10554207116365433,
      "learning_rate": 0.0001586423118503638,
      "loss": 1.1114,
      "step": 1672
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.15037661790847778,
      "learning_rate": 0.0001585918202946334,
      "loss": 1.1083,
      "step": 1673
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.15390411019325256,
      "learning_rate": 0.00015854130598251512,
      "loss": 1.1669,
      "step": 1674
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.11049924045801163,
      "learning_rate": 0.0001584907689336282,
      "loss": 1.1585,
      "step": 1675
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.11888211220502853,
      "learning_rate": 0.0001584402091676006,
      "loss": 1.2321,
      "step": 1676
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1601606011390686,
      "learning_rate": 0.00015838962670406916,
      "loss": 1.1914,
      "step": 1677
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.12408497929573059,
      "learning_rate": 0.00015833902156267956,
      "loss": 1.1702,
      "step": 1678
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.12359727174043655,
      "learning_rate": 0.00015828839376308618,
      "loss": 1.1255,
      "step": 1679
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.13961875438690186,
      "learning_rate": 0.00015823774332495235,
      "loss": 1.1697,
      "step": 1680
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.1547592729330063,
      "learning_rate": 0.0001581870702679501,
      "loss": 1.1807,
      "step": 1681
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.13319851458072662,
      "learning_rate": 0.0001581363746117602,
      "loss": 1.077,
      "step": 1682
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.10309629142284393,
      "learning_rate": 0.00015808565637607237,
      "loss": 1.0999,
      "step": 1683
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.13052968680858612,
      "learning_rate": 0.00015803491558058488,
      "loss": 1.14,
      "step": 1684
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.09577590972185135,
      "learning_rate": 0.00015798415224500492,
      "loss": 1.0444,
      "step": 1685
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.12723176181316376,
      "learning_rate": 0.00015793336638904838,
      "loss": 1.1961,
      "step": 1686
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.10387569665908813,
      "learning_rate": 0.0001578825580324399,
      "loss": 1.0472,
      "step": 1687
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.13047683238983154,
      "learning_rate": 0.0001578317271949129,
      "loss": 1.1609,
      "step": 1688
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.12225230783224106,
      "learning_rate": 0.00015778087389620938,
      "loss": 1.0854,
      "step": 1689
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.13382606208324432,
      "learning_rate": 0.00015772999815608028,
      "loss": 1.0628,
      "step": 1690
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.12056230753660202,
      "learning_rate": 0.00015767909999428513,
      "loss": 1.167,
      "step": 1691
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.23597656190395355,
      "learning_rate": 0.00015762817943059217,
      "loss": 1.2056,
      "step": 1692
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.11973971873521805,
      "learning_rate": 0.00015757723648477837,
      "loss": 1.1154,
      "step": 1693
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.10275455564260483,
      "learning_rate": 0.0001575262711766294,
      "loss": 1.1166,
      "step": 1694
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.14231620728969574,
      "learning_rate": 0.00015747528352593956,
      "loss": 1.2,
      "step": 1695
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.13756804168224335,
      "learning_rate": 0.0001574242735525119,
      "loss": 1.1458,
      "step": 1696
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1772712767124176,
      "learning_rate": 0.00015737324127615806,
      "loss": 1.2491,
      "step": 1697
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1378726214170456,
      "learning_rate": 0.00015732218671669844,
      "loss": 1.1872,
      "step": 1698
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.14357145130634308,
      "learning_rate": 0.00015727110989396202,
      "loss": 1.0129,
      "step": 1699
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.16498640179634094,
      "learning_rate": 0.00015722001082778646,
      "loss": 1.1423,
      "step": 1700
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.13786764442920685,
      "learning_rate": 0.00015716888953801804,
      "loss": 1.1467,
      "step": 1701
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.17314563691616058,
      "learning_rate": 0.00015711774604451167,
      "loss": 1.1285,
      "step": 1702
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.11548225581645966,
      "learning_rate": 0.0001570665803671309,
      "loss": 1.1752,
      "step": 1703
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.10307903587818146,
      "learning_rate": 0.00015701539252574792,
      "loss": 1.1047,
      "step": 1704
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.13165180385112762,
      "learning_rate": 0.00015696418254024344,
      "loss": 1.0707,
      "step": 1705
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.14903388917446136,
      "learning_rate": 0.00015691295043050688,
      "loss": 1.1812,
      "step": 1706
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.11531300842761993,
      "learning_rate": 0.0001568616962164362,
      "loss": 1.219,
      "step": 1707
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.10394908487796783,
      "learning_rate": 0.0001568104199179379,
      "loss": 1.1162,
      "step": 1708
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.14028772711753845,
      "learning_rate": 0.00015675912155492712,
      "loss": 1.1664,
      "step": 1709
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.11324262619018555,
      "learning_rate": 0.00015670780114732756,
      "loss": 1.1051,
      "step": 1710
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.12212210893630981,
      "learning_rate": 0.00015665645871507151,
      "loss": 1.0592,
      "step": 1711
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1250733733177185,
      "learning_rate": 0.00015660509427809974,
      "loss": 1.1305,
      "step": 1712
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.07854495197534561,
      "learning_rate": 0.0001565537078563616,
      "loss": 1.1807,
      "step": 1713
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.10706120729446411,
      "learning_rate": 0.000156502299469815,
      "loss": 1.1565,
      "step": 1714
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.11902441084384918,
      "learning_rate": 0.00015645086913842636,
      "loss": 1.1943,
      "step": 1715
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.12750212848186493,
      "learning_rate": 0.00015639941688217065,
      "loss": 1.1854,
      "step": 1716
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.10871671140193939,
      "learning_rate": 0.00015634794272103127,
      "loss": 1.1435,
      "step": 1717
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.14880502223968506,
      "learning_rate": 0.0001562964466750003,
      "loss": 1.175,
      "step": 1718
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1627848893404007,
      "learning_rate": 0.0001562449287640781,
      "loss": 1.1917,
      "step": 1719
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.15309424698352814,
      "learning_rate": 0.00015619338900827367,
      "loss": 1.099,
      "step": 1720
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.12216850370168686,
      "learning_rate": 0.00015614182742760448,
      "loss": 1.1706,
      "step": 1721
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.12746180593967438,
      "learning_rate": 0.00015609024404209643,
      "loss": 1.1834,
      "step": 1722
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1308891922235489,
      "learning_rate": 0.00015603863887178393,
      "loss": 1.2256,
      "step": 1723
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.14137287437915802,
      "learning_rate": 0.00015598701193670982,
      "loss": 1.1067,
      "step": 1724
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.10840347409248352,
      "learning_rate": 0.0001559353632569254,
      "loss": 1.1658,
      "step": 1725
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1199110820889473,
      "learning_rate": 0.00015588369285249047,
      "loss": 1.187,
      "step": 1726
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.14508295059204102,
      "learning_rate": 0.00015583200074347316,
      "loss": 1.0967,
      "step": 1727
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.11529313772916794,
      "learning_rate": 0.0001557802869499501,
      "loss": 1.1409,
      "step": 1728
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.15027004480361938,
      "learning_rate": 0.00015572855149200638,
      "loss": 1.1768,
      "step": 1729
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.16301485896110535,
      "learning_rate": 0.0001556767943897354,
      "loss": 1.1478,
      "step": 1730
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1456843465566635,
      "learning_rate": 0.00015562501566323907,
      "loss": 1.1524,
      "step": 1731
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1253364235162735,
      "learning_rate": 0.0001555732153326276,
      "loss": 1.1195,
      "step": 1732
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1435948610305786,
      "learning_rate": 0.00015552139341801967,
      "loss": 1.0382,
      "step": 1733
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.11362790316343307,
      "learning_rate": 0.00015546954993954227,
      "loss": 1.1522,
      "step": 1734
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.146712064743042,
      "learning_rate": 0.00015541768491733092,
      "loss": 1.1252,
      "step": 1735
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.10069835186004639,
      "learning_rate": 0.00015536579837152926,
      "loss": 1.0611,
      "step": 1736
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.11298953741788864,
      "learning_rate": 0.00015531389032228955,
      "loss": 1.1197,
      "step": 1737
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.12217997759580612,
      "learning_rate": 0.00015526196078977217,
      "loss": 1.127,
      "step": 1738
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.10973166674375534,
      "learning_rate": 0.00015521000979414602,
      "loss": 1.0655,
      "step": 1739
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1495875120162964,
      "learning_rate": 0.00015515803735558826,
      "loss": 1.2422,
      "step": 1740
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.15007765591144562,
      "learning_rate": 0.00015510604349428436,
      "loss": 1.1591,
      "step": 1741
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.10734861344099045,
      "learning_rate": 0.00015505402823042818,
      "loss": 1.1437,
      "step": 1742
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13395027816295624,
      "learning_rate": 0.00015500199158422178,
      "loss": 1.1992,
      "step": 1743
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.10562297701835632,
      "learning_rate": 0.0001549499335758757,
      "loss": 1.1596,
      "step": 1744
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1235433965921402,
      "learning_rate": 0.00015489785422560857,
      "loss": 1.0993,
      "step": 1745
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1195843443274498,
      "learning_rate": 0.00015484575355364743,
      "loss": 1.0878,
      "step": 1746
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1292940080165863,
      "learning_rate": 0.00015479363158022764,
      "loss": 1.2005,
      "step": 1747
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1376148760318756,
      "learning_rate": 0.0001547414883255927,
      "loss": 1.1355,
      "step": 1748
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.09842343628406525,
      "learning_rate": 0.0001546893238099945,
      "loss": 1.1561,
      "step": 1749
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1298367828130722,
      "learning_rate": 0.0001546371380536931,
      "loss": 1.1288,
      "step": 1750
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1482243835926056,
      "learning_rate": 0.00015458493107695686,
      "loss": 1.1606,
      "step": 1751
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.14012926816940308,
      "learning_rate": 0.00015453270290006238,
      "loss": 1.3042,
      "step": 1752
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.11902008950710297,
      "learning_rate": 0.00015448045354329447,
      "loss": 1.088,
      "step": 1753
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.12221182137727737,
      "learning_rate": 0.00015442818302694618,
      "loss": 1.1175,
      "step": 1754
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13082818686962128,
      "learning_rate": 0.0001543758913713188,
      "loss": 1.2123,
      "step": 1755
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1537473350763321,
      "learning_rate": 0.00015432357859672177,
      "loss": 1.209,
      "step": 1756
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.12102311104536057,
      "learning_rate": 0.00015427124472347278,
      "loss": 1.201,
      "step": 1757
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.10943295061588287,
      "learning_rate": 0.0001542188897718977,
      "loss": 1.1952,
      "step": 1758
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.15755842626094818,
      "learning_rate": 0.00015416651376233063,
      "loss": 1.1581,
      "step": 1759
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1363496035337448,
      "learning_rate": 0.00015411411671511377,
      "loss": 1.1004,
      "step": 1760
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.10768219083547592,
      "learning_rate": 0.00015406169865059749,
      "loss": 1.1491,
      "step": 1761
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.11993807554244995,
      "learning_rate": 0.00015400925958914042,
      "loss": 1.1938,
      "step": 1762
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13427366316318512,
      "learning_rate": 0.00015395679955110925,
      "loss": 1.1223,
      "step": 1763
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13078995048999786,
      "learning_rate": 0.00015390431855687896,
      "loss": 1.1208,
      "step": 1764
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1448216289281845,
      "learning_rate": 0.00015385181662683244,
      "loss": 1.169,
      "step": 1765
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13500584661960602,
      "learning_rate": 0.00015379929378136087,
      "loss": 1.1586,
      "step": 1766
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.11047647148370743,
      "learning_rate": 0.00015374675004086355,
      "loss": 1.1286,
      "step": 1767
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.15553277730941772,
      "learning_rate": 0.00015369418542574782,
      "loss": 1.2205,
      "step": 1768
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.12137002497911453,
      "learning_rate": 0.00015364159995642917,
      "loss": 1.135,
      "step": 1769
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.11878468096256256,
      "learning_rate": 0.00015358899365333124,
      "loss": 1.1345,
      "step": 1770
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.12524281442165375,
      "learning_rate": 0.00015353636653688563,
      "loss": 1.122,
      "step": 1771
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.12499994784593582,
      "learning_rate": 0.0001534837186275322,
      "loss": 1.1476,
      "step": 1772
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.2054329365491867,
      "learning_rate": 0.00015343104994571875,
      "loss": 1.0836,
      "step": 1773
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.09046951681375504,
      "learning_rate": 0.0001533783605119012,
      "loss": 1.1219,
      "step": 1774
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1281549483537674,
      "learning_rate": 0.00015332565034654344,
      "loss": 1.2119,
      "step": 1775
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.11029399186372757,
      "learning_rate": 0.00015327291947011762,
      "loss": 1.2011,
      "step": 1776
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.10453195124864578,
      "learning_rate": 0.00015322016790310372,
      "loss": 1.1654,
      "step": 1777
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13024407625198364,
      "learning_rate": 0.00015316739566598986,
      "loss": 1.1886,
      "step": 1778
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13169990479946136,
      "learning_rate": 0.00015311460277927217,
      "loss": 1.1082,
      "step": 1779
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.12379168719053268,
      "learning_rate": 0.0001530617892634548,
      "loss": 1.0638,
      "step": 1780
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.10757247358560562,
      "learning_rate": 0.0001530089551390499,
      "loss": 1.0748,
      "step": 1781
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.09000054746866226,
      "learning_rate": 0.0001529561004265777,
      "loss": 1.1339,
      "step": 1782
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.12435159087181091,
      "learning_rate": 0.00015290322514656626,
      "loss": 1.0563,
      "step": 1783
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.09496708959341049,
      "learning_rate": 0.00015285032931955177,
      "loss": 1.1164,
      "step": 1784
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13473650813102722,
      "learning_rate": 0.0001527974129660784,
      "loss": 1.1083,
      "step": 1785
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.12117791175842285,
      "learning_rate": 0.0001527444761066982,
      "loss": 1.0944,
      "step": 1786
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.11469109356403351,
      "learning_rate": 0.00015269151876197125,
      "loss": 1.1142,
      "step": 1787
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.12138448655605316,
      "learning_rate": 0.00015263854095246557,
      "loss": 1.0825,
      "step": 1788
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13748134672641754,
      "learning_rate": 0.00015258554269875717,
      "loss": 1.0955,
      "step": 1789
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.14604851603507996,
      "learning_rate": 0.00015253252402142988,
      "loss": 1.0895,
      "step": 1790
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.13862721621990204,
      "learning_rate": 0.00015247948494107565,
      "loss": 1.128,
      "step": 1791
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1347518414258957,
      "learning_rate": 0.00015242642547829417,
      "loss": 1.0956,
      "step": 1792
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.125281423330307,
      "learning_rate": 0.0001523733456536931,
      "loss": 1.1382,
      "step": 1793
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.17020346224308014,
      "learning_rate": 0.00015232024548788813,
      "loss": 1.1716,
      "step": 1794
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.148721382021904,
      "learning_rate": 0.00015226712500150268,
      "loss": 1.1265,
      "step": 1795
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.149869903922081,
      "learning_rate": 0.00015221398421516816,
      "loss": 1.1832,
      "step": 1796
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.13809417188167572,
      "learning_rate": 0.0001521608231495238,
      "loss": 1.1456,
      "step": 1797
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12363449484109879,
      "learning_rate": 0.0001521076418252168,
      "loss": 1.1561,
      "step": 1798
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12657707929611206,
      "learning_rate": 0.00015205444026290216,
      "loss": 1.1685,
      "step": 1799
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.14082394540309906,
      "learning_rate": 0.00015200121848324275,
      "loss": 1.0952,
      "step": 1800
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12948375940322876,
      "learning_rate": 0.00015194797650690926,
      "loss": 1.1752,
      "step": 1801
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.124544657766819,
      "learning_rate": 0.0001518947143545803,
      "loss": 1.1047,
      "step": 1802
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11809641867876053,
      "learning_rate": 0.0001518414320469423,
      "loss": 1.1384,
      "step": 1803
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.16011102497577667,
      "learning_rate": 0.00015178812960468945,
      "loss": 1.1528,
      "step": 1804
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12365131080150604,
      "learning_rate": 0.0001517348070485238,
      "loss": 1.1294,
      "step": 1805
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.15388205647468567,
      "learning_rate": 0.00015168146439915525,
      "loss": 1.2042,
      "step": 1806
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11523950845003128,
      "learning_rate": 0.00015162810167730143,
      "loss": 1.0831,
      "step": 1807
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11957983672618866,
      "learning_rate": 0.00015157471890368785,
      "loss": 1.1861,
      "step": 1808
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12794260680675507,
      "learning_rate": 0.0001515213160990477,
      "loss": 1.193,
      "step": 1809
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.14271734654903412,
      "learning_rate": 0.00015146789328412212,
      "loss": 1.0825,
      "step": 1810
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.10542784631252289,
      "learning_rate": 0.00015141445047965984,
      "loss": 1.0902,
      "step": 1811
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.13587680459022522,
      "learning_rate": 0.0001513609877064174,
      "loss": 1.136,
      "step": 1812
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11955380439758301,
      "learning_rate": 0.0001513075049851592,
      "loss": 1.1814,
      "step": 1813
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.18435966968536377,
      "learning_rate": 0.0001512540023366573,
      "loss": 1.1092,
      "step": 1814
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.14707142114639282,
      "learning_rate": 0.00015120047978169144,
      "loss": 1.0934,
      "step": 1815
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1424715220928192,
      "learning_rate": 0.00015114693734104927,
      "loss": 1.1206,
      "step": 1816
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.09493036568164825,
      "learning_rate": 0.00015109337503552595,
      "loss": 1.1714,
      "step": 1817
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11911047250032425,
      "learning_rate": 0.00015103979288592453,
      "loss": 1.0711,
      "step": 1818
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.15567700564861298,
      "learning_rate": 0.0001509861909130557,
      "loss": 1.1147,
      "step": 1819
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.10658947378396988,
      "learning_rate": 0.00015093256913773786,
      "loss": 1.087,
      "step": 1820
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.13488833606243134,
      "learning_rate": 0.000150878927580797,
      "loss": 1.1356,
      "step": 1821
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.14417381584644318,
      "learning_rate": 0.00015082526626306695,
      "loss": 1.2281,
      "step": 1822
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12065943330526352,
      "learning_rate": 0.0001507715852053892,
      "loss": 1.2163,
      "step": 1823
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1251625269651413,
      "learning_rate": 0.00015071788442861276,
      "loss": 1.0693,
      "step": 1824
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1078992411494255,
      "learning_rate": 0.00015066416395359444,
      "loss": 1.1976,
      "step": 1825
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.17007358372211456,
      "learning_rate": 0.00015061042380119864,
      "loss": 1.2307,
      "step": 1826
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.10376029461622238,
      "learning_rate": 0.0001505566639922974,
      "loss": 1.2593,
      "step": 1827
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11600358039140701,
      "learning_rate": 0.00015050288454777046,
      "loss": 1.0992,
      "step": 1828
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.161931112408638,
      "learning_rate": 0.0001504490854885051,
      "loss": 1.1632,
      "step": 1829
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1488901823759079,
      "learning_rate": 0.00015039526683539625,
      "loss": 1.1497,
      "step": 1830
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12617997825145721,
      "learning_rate": 0.0001503414286093465,
      "loss": 1.0794,
      "step": 1831
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11766906827688217,
      "learning_rate": 0.00015028757083126592,
      "loss": 1.1949,
      "step": 1832
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.14150752127170563,
      "learning_rate": 0.00015023369352207229,
      "loss": 1.1677,
      "step": 1833
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.11629250645637512,
      "learning_rate": 0.00015017979670269095,
      "loss": 1.1093,
      "step": 1834
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1302558332681656,
      "learning_rate": 0.0001501258803940548,
      "loss": 1.1924,
      "step": 1835
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.10650352388620377,
      "learning_rate": 0.0001500719446171043,
      "loss": 1.1433,
      "step": 1836
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1648806780576706,
      "learning_rate": 0.00015001798939278752,
      "loss": 1.2045,
      "step": 1837
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1283397227525711,
      "learning_rate": 0.00014996401474205997,
      "loss": 1.2166,
      "step": 1838
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1222076416015625,
      "learning_rate": 0.00014991002068588484,
      "loss": 1.1342,
      "step": 1839
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12195917963981628,
      "learning_rate": 0.0001498560072452328,
      "loss": 1.1191,
      "step": 1840
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.14561673998832703,
      "learning_rate": 0.00014980197444108205,
      "loss": 1.2262,
      "step": 1841
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1277858018875122,
      "learning_rate": 0.00014974792229441826,
      "loss": 1.2199,
      "step": 1842
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12258052080869675,
      "learning_rate": 0.00014969385082623472,
      "loss": 1.1449,
      "step": 1843
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.15511752665042877,
      "learning_rate": 0.00014963976005753215,
      "loss": 1.1122,
      "step": 1844
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.12786231935024261,
      "learning_rate": 0.00014958565000931876,
      "loss": 1.1038,
      "step": 1845
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.14985772967338562,
      "learning_rate": 0.00014953152070261026,
      "loss": 1.1795,
      "step": 1846
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1209183782339096,
      "learning_rate": 0.0001494773721584299,
      "loss": 1.175,
      "step": 1847
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12308932095766068,
      "learning_rate": 0.0001494232043978083,
      "loss": 1.1992,
      "step": 1848
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.11509066820144653,
      "learning_rate": 0.00014936901744178367,
      "loss": 1.2145,
      "step": 1849
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.10604944825172424,
      "learning_rate": 0.00014931481131140147,
      "loss": 1.1568,
      "step": 1850
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.10772743821144104,
      "learning_rate": 0.00014926058602771484,
      "loss": 1.1623,
      "step": 1851
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.11908062547445297,
      "learning_rate": 0.00014920634161178425,
      "loss": 1.1738,
      "step": 1852
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12726634740829468,
      "learning_rate": 0.00014915207808467756,
      "loss": 1.2281,
      "step": 1853
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.15437252819538116,
      "learning_rate": 0.0001490977954674701,
      "loss": 1.2044,
      "step": 1854
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1712338626384735,
      "learning_rate": 0.00014904349378124467,
      "loss": 1.1449,
      "step": 1855
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12339615076780319,
      "learning_rate": 0.0001489891730470914,
      "loss": 1.1434,
      "step": 1856
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.14506807923316956,
      "learning_rate": 0.00014893483328610777,
      "loss": 1.1785,
      "step": 1857
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13917265832424164,
      "learning_rate": 0.0001488804745193988,
      "loss": 1.0485,
      "step": 1858
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.14300934970378876,
      "learning_rate": 0.00014882609676807675,
      "loss": 1.1614,
      "step": 1859
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1567572057247162,
      "learning_rate": 0.00014877170005326136,
      "loss": 1.1261,
      "step": 1860
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12624302506446838,
      "learning_rate": 0.00014871728439607966,
      "loss": 1.1499,
      "step": 1861
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1413816511631012,
      "learning_rate": 0.00014866284981766606,
      "loss": 1.2281,
      "step": 1862
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.11266129463911057,
      "learning_rate": 0.00014860839633916236,
      "loss": 1.1017,
      "step": 1863
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.15663382411003113,
      "learning_rate": 0.0001485539239817176,
      "loss": 1.1327,
      "step": 1864
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13321727514266968,
      "learning_rate": 0.0001484994327664883,
      "loss": 1.1742,
      "step": 1865
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12131322920322418,
      "learning_rate": 0.0001484449227146381,
      "loss": 1.1768,
      "step": 1866
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.11533010751008987,
      "learning_rate": 0.0001483903938473382,
      "loss": 1.1173,
      "step": 1867
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.08990739285945892,
      "learning_rate": 0.00014833584618576695,
      "loss": 1.0572,
      "step": 1868
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12184695154428482,
      "learning_rate": 0.00014828127975111,
      "loss": 1.2295,
      "step": 1869
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.11232170462608337,
      "learning_rate": 0.0001482266945645603,
      "loss": 1.1772,
      "step": 1870
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13341142237186432,
      "learning_rate": 0.00014817209064731817,
      "loss": 1.1909,
      "step": 1871
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12435612082481384,
      "learning_rate": 0.00014811746802059113,
      "loss": 1.1113,
      "step": 1872
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.17336216568946838,
      "learning_rate": 0.000148062826705594,
      "loss": 1.1116,
      "step": 1873
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1421254724264145,
      "learning_rate": 0.00014800816672354877,
      "loss": 1.1358,
      "step": 1874
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.10679803043603897,
      "learning_rate": 0.00014795348809568476,
      "loss": 1.1152,
      "step": 1875
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12673167884349823,
      "learning_rate": 0.00014789879084323857,
      "loss": 1.1486,
      "step": 1876
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.15120945870876312,
      "learning_rate": 0.00014784407498745394,
      "loss": 1.1491,
      "step": 1877
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.162373885512352,
      "learning_rate": 0.0001477893405495819,
      "loss": 1.1504,
      "step": 1878
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1504976898431778,
      "learning_rate": 0.00014773458755088068,
      "loss": 1.0699,
      "step": 1879
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.11421702057123184,
      "learning_rate": 0.00014767981601261567,
      "loss": 1.1657,
      "step": 1880
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12368762493133545,
      "learning_rate": 0.00014762502595605956,
      "loss": 1.2081,
      "step": 1881
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12744393944740295,
      "learning_rate": 0.00014757021740249214,
      "loss": 1.1836,
      "step": 1882
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.10182986408472061,
      "learning_rate": 0.00014751539037320043,
      "loss": 1.039,
      "step": 1883
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13908879458904266,
      "learning_rate": 0.00014746054488947862,
      "loss": 1.2131,
      "step": 1884
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1319282352924347,
      "learning_rate": 0.00014740568097262812,
      "loss": 1.1758,
      "step": 1885
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.18841414153575897,
      "learning_rate": 0.00014735079864395729,
      "loss": 1.1098,
      "step": 1886
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.113658107817173,
      "learning_rate": 0.00014729589792478192,
      "loss": 1.1287,
      "step": 1887
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.15880362689495087,
      "learning_rate": 0.00014724097883642482,
      "loss": 1.1856,
      "step": 1888
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1454492211341858,
      "learning_rate": 0.00014718604140021588,
      "loss": 1.0601,
      "step": 1889
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.1331673115491867,
      "learning_rate": 0.0001471310856374922,
      "loss": 1.1621,
      "step": 1890
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13103727996349335,
      "learning_rate": 0.0001470761115695979,
      "loss": 1.1183,
      "step": 1891
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.10028551518917084,
      "learning_rate": 0.00014702111921788437,
      "loss": 1.1475,
      "step": 1892
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13771052658557892,
      "learning_rate": 0.00014696610860370996,
      "loss": 1.156,
      "step": 1893
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.18962733447551727,
      "learning_rate": 0.00014691107974844014,
      "loss": 1.108,
      "step": 1894
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13492996990680695,
      "learning_rate": 0.0001468560326734475,
      "loss": 1.1488,
      "step": 1895
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.12778440117835999,
      "learning_rate": 0.00014680096740011172,
      "loss": 1.1547,
      "step": 1896
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.10388492792844772,
      "learning_rate": 0.00014674588394981947,
      "loss": 1.1053,
      "step": 1897
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.12216372787952423,
      "learning_rate": 0.00014669078234396453,
      "loss": 1.1634,
      "step": 1898
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.12342170625925064,
      "learning_rate": 0.00014663566260394775,
      "loss": 1.1198,
      "step": 1899
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13942089676856995,
      "learning_rate": 0.00014658052475117702,
      "loss": 1.1209,
      "step": 1900
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.20593951642513275,
      "learning_rate": 0.0001465253688070672,
      "loss": 1.1713,
      "step": 1901
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13371625542640686,
      "learning_rate": 0.00014647019479304028,
      "loss": 1.0939,
      "step": 1902
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.18048664927482605,
      "learning_rate": 0.00014641500273052514,
      "loss": 1.1531,
      "step": 1903
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13993196189403534,
      "learning_rate": 0.0001463597926409578,
      "loss": 1.1763,
      "step": 1904
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.14304153621196747,
      "learning_rate": 0.0001463045645457812,
      "loss": 1.1808,
      "step": 1905
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.14477460086345673,
      "learning_rate": 0.0001462493184664453,
      "loss": 1.187,
      "step": 1906
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13996070623397827,
      "learning_rate": 0.000146194054424407,
      "loss": 1.1108,
      "step": 1907
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1357588768005371,
      "learning_rate": 0.00014613877244113032,
      "loss": 1.1096,
      "step": 1908
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.14302358031272888,
      "learning_rate": 0.00014608347253808606,
      "loss": 1.222,
      "step": 1909
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1295420378446579,
      "learning_rate": 0.0001460281547367521,
      "loss": 1.155,
      "step": 1910
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13043150305747986,
      "learning_rate": 0.0001459728190586132,
      "loss": 1.1651,
      "step": 1911
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13847218453884125,
      "learning_rate": 0.00014591746552516108,
      "loss": 1.2001,
      "step": 1912
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.11023901402950287,
      "learning_rate": 0.0001458620941578945,
      "loss": 1.133,
      "step": 1913
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1313227266073227,
      "learning_rate": 0.00014580670497831903,
      "loss": 1.1936,
      "step": 1914
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.10304143279790878,
      "learning_rate": 0.00014575129800794718,
      "loss": 1.1158,
      "step": 1915
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13406261801719666,
      "learning_rate": 0.00014569587326829835,
      "loss": 1.2015,
      "step": 1916
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.14598830044269562,
      "learning_rate": 0.0001456404307808989,
      "loss": 1.1654,
      "step": 1917
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1112588569521904,
      "learning_rate": 0.00014558497056728204,
      "loss": 1.0795,
      "step": 1918
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1407684087753296,
      "learning_rate": 0.00014552949264898794,
      "loss": 1.1212,
      "step": 1919
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.15352100133895874,
      "learning_rate": 0.00014547399704756348,
      "loss": 1.151,
      "step": 1920
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13194423913955688,
      "learning_rate": 0.00014541848378456256,
      "loss": 1.182,
      "step": 1921
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.14242687821388245,
      "learning_rate": 0.00014536295288154592,
      "loss": 1.0935,
      "step": 1922
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.12536852061748505,
      "learning_rate": 0.0001453074043600811,
      "loss": 1.0492,
      "step": 1923
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.10591185837984085,
      "learning_rate": 0.00014525183824174248,
      "loss": 1.2182,
      "step": 1924
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.11347635835409164,
      "learning_rate": 0.00014519625454811135,
      "loss": 1.165,
      "step": 1925
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.11378409713506699,
      "learning_rate": 0.00014514065330077576,
      "loss": 1.1204,
      "step": 1926
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.16188538074493408,
      "learning_rate": 0.00014508503452133052,
      "loss": 1.0869,
      "step": 1927
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1350652277469635,
      "learning_rate": 0.00014502939823137745,
      "loss": 1.1548,
      "step": 1928
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.11640197038650513,
      "learning_rate": 0.00014497374445252497,
      "loss": 1.1546,
      "step": 1929
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.15174321830272675,
      "learning_rate": 0.00014491807320638834,
      "loss": 1.0594,
      "step": 1930
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1480076164007187,
      "learning_rate": 0.0001448623845145897,
      "loss": 1.1777,
      "step": 1931
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13455988466739655,
      "learning_rate": 0.00014480667839875786,
      "loss": 1.2137,
      "step": 1932
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.10894989222288132,
      "learning_rate": 0.00014475095488052843,
      "loss": 1.1436,
      "step": 1933
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13975659012794495,
      "learning_rate": 0.0001446952139815438,
      "loss": 1.0589,
      "step": 1934
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.15277092158794403,
      "learning_rate": 0.00014463945572345307,
      "loss": 1.1631,
      "step": 1935
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.10714533179998398,
      "learning_rate": 0.00014458368012791213,
      "loss": 1.2273,
      "step": 1936
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1476077437400818,
      "learning_rate": 0.00014452788721658355,
      "loss": 1.0585,
      "step": 1937
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.12188824266195297,
      "learning_rate": 0.00014447207701113668,
      "loss": 1.0715,
      "step": 1938
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.09883607923984528,
      "learning_rate": 0.00014441624953324755,
      "loss": 1.2111,
      "step": 1939
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.15379509329795837,
      "learning_rate": 0.00014436040480459892,
      "loss": 1.137,
      "step": 1940
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.11753641813993454,
      "learning_rate": 0.0001443045428468802,
      "loss": 1.1747,
      "step": 1941
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1511080265045166,
      "learning_rate": 0.0001442486636817876,
      "loss": 1.2122,
      "step": 1942
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.16177278757095337,
      "learning_rate": 0.00014419276733102388,
      "loss": 1.1208,
      "step": 1943
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.16989944875240326,
      "learning_rate": 0.00014413685381629856,
      "loss": 1.1574,
      "step": 1944
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.10415186733007431,
      "learning_rate": 0.0001440809231593278,
      "loss": 1.127,
      "step": 1945
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.14299753308296204,
      "learning_rate": 0.00014402497538183444,
      "loss": 1.0972,
      "step": 1946
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13852828741073608,
      "learning_rate": 0.00014396901050554793,
      "loss": 1.1131,
      "step": 1947
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.11999713629484177,
      "learning_rate": 0.0001439130285522044,
      "loss": 1.1156,
      "step": 1948
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.1025676429271698,
      "learning_rate": 0.0001438570295435466,
      "loss": 1.1634,
      "step": 1949
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1258241981267929,
      "learning_rate": 0.00014380101350132388,
      "loss": 1.1994,
      "step": 1950
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10830667614936829,
      "learning_rate": 0.00014374498044729227,
      "loss": 1.132,
      "step": 1951
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11750224232673645,
      "learning_rate": 0.00014368893040321427,
      "loss": 1.1473,
      "step": 1952
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12033667415380478,
      "learning_rate": 0.00014363286339085914,
      "loss": 1.0473,
      "step": 1953
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.16292725503444672,
      "learning_rate": 0.0001435767794320027,
      "loss": 1.213,
      "step": 1954
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10896769165992737,
      "learning_rate": 0.00014352067854842723,
      "loss": 1.1094,
      "step": 1955
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.13429225981235504,
      "learning_rate": 0.0001434645607619217,
      "loss": 1.2002,
      "step": 1956
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1637416034936905,
      "learning_rate": 0.0001434084260942816,
      "loss": 1.1294,
      "step": 1957
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.09355669468641281,
      "learning_rate": 0.000143352274567309,
      "loss": 1.087,
      "step": 1958
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10112952440977097,
      "learning_rate": 0.00014329610620281253,
      "loss": 1.1057,
      "step": 1959
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.14689409732818604,
      "learning_rate": 0.00014323992102260733,
      "loss": 1.1392,
      "step": 1960
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.16043992340564728,
      "learning_rate": 0.00014318371904851503,
      "loss": 1.1306,
      "step": 1961
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10764963179826736,
      "learning_rate": 0.00014312750030236383,
      "loss": 1.209,
      "step": 1962
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12133940309286118,
      "learning_rate": 0.0001430712648059885,
      "loss": 1.0708,
      "step": 1963
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.13435018062591553,
      "learning_rate": 0.00014301501258123025,
      "loss": 1.1121,
      "step": 1964
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1402486264705658,
      "learning_rate": 0.0001429587436499367,
      "loss": 1.0896,
      "step": 1965
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.14691080152988434,
      "learning_rate": 0.0001429024580339622,
      "loss": 1.1301,
      "step": 1966
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.128518745303154,
      "learning_rate": 0.00014284615575516738,
      "loss": 1.1891,
      "step": 1967
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10371225327253342,
      "learning_rate": 0.00014278983683541934,
      "loss": 1.1206,
      "step": 1968
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11523692309856415,
      "learning_rate": 0.00014273350129659172,
      "loss": 1.1735,
      "step": 1969
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.15106497704982758,
      "learning_rate": 0.00014267714916056464,
      "loss": 1.2599,
      "step": 1970
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1044711172580719,
      "learning_rate": 0.0001426207804492246,
      "loss": 1.2146,
      "step": 1971
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12189260125160217,
      "learning_rate": 0.00014256439518446455,
      "loss": 1.0775,
      "step": 1972
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11257151514291763,
      "learning_rate": 0.00014250799338818388,
      "loss": 1.1258,
      "step": 1973
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12427462637424469,
      "learning_rate": 0.00014245157508228839,
      "loss": 1.1692,
      "step": 1974
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.18095479905605316,
      "learning_rate": 0.00014239514028869033,
      "loss": 1.1924,
      "step": 1975
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11171863973140717,
      "learning_rate": 0.00014233868902930826,
      "loss": 1.1529,
      "step": 1976
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12951019406318665,
      "learning_rate": 0.0001422822213260673,
      "loss": 1.1354,
      "step": 1977
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.13598938286304474,
      "learning_rate": 0.00014222573720089874,
      "loss": 1.0454,
      "step": 1978
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.15544407069683075,
      "learning_rate": 0.00014216923667574043,
      "loss": 1.1758,
      "step": 1979
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.13198311626911163,
      "learning_rate": 0.00014211271977253653,
      "loss": 1.1327,
      "step": 1980
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1398787796497345,
      "learning_rate": 0.00014205618651323752,
      "loss": 1.0806,
      "step": 1981
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11922745406627655,
      "learning_rate": 0.00014199963691980026,
      "loss": 1.0557,
      "step": 1982
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11732225865125656,
      "learning_rate": 0.00014194307101418805,
      "loss": 1.2209,
      "step": 1983
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11124857515096664,
      "learning_rate": 0.00014188648881837033,
      "loss": 1.1275,
      "step": 1984
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10288741439580917,
      "learning_rate": 0.00014182989035432298,
      "loss": 1.1793,
      "step": 1985
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.14518317580223083,
      "learning_rate": 0.00014177327564402826,
      "loss": 1.0727,
      "step": 1986
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1108485758304596,
      "learning_rate": 0.00014171664470947465,
      "loss": 1.0562,
      "step": 1987
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1490449756383896,
      "learning_rate": 0.0001416599975726569,
      "loss": 1.1019,
      "step": 1988
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10886706411838531,
      "learning_rate": 0.00014160333425557615,
      "loss": 1.195,
      "step": 1989
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1290430873632431,
      "learning_rate": 0.00014154665478023976,
      "loss": 1.1279,
      "step": 1990
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1199612021446228,
      "learning_rate": 0.0001414899591686614,
      "loss": 1.218,
      "step": 1991
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11188086122274399,
      "learning_rate": 0.000141433247442861,
      "loss": 1.1645,
      "step": 1992
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12187888473272324,
      "learning_rate": 0.0001413765196248647,
      "loss": 1.1234,
      "step": 1993
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12547385692596436,
      "learning_rate": 0.00014131977573670498,
      "loss": 1.1014,
      "step": 1994
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10714356601238251,
      "learning_rate": 0.0001412630158004205,
      "loss": 1.2182,
      "step": 1995
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11527112871408463,
      "learning_rate": 0.00014120623983805616,
      "loss": 1.1022,
      "step": 1996
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1490398794412613,
      "learning_rate": 0.00014114944787166307,
      "loss": 1.1678,
      "step": 1997
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.07947546243667603,
      "learning_rate": 0.00014109263992329857,
      "loss": 1.2045,
      "step": 1998
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.11817868053913116,
      "learning_rate": 0.0001410358160150263,
      "loss": 1.151,
      "step": 1999
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.13336190581321716,
      "learning_rate": 0.0001409789761689159,
      "loss": 1.1455,
      "step": 2000
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.12902678549289703,
      "learning_rate": 0.00014092212040704336,
      "loss": 1.1455,
      "step": 2001
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.15024402737617493,
      "learning_rate": 0.0001408652487514908,
      "loss": 1.1372,
      "step": 2002
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.12081260234117508,
      "learning_rate": 0.0001408083612243465,
      "loss": 1.2106,
      "step": 2003
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.12936916947364807,
      "learning_rate": 0.00014075145784770495,
      "loss": 1.2424,
      "step": 2004
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1285601109266281,
      "learning_rate": 0.00014069453864366677,
      "loss": 1.12,
      "step": 2005
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1358678936958313,
      "learning_rate": 0.00014063760363433866,
      "loss": 1.2141,
      "step": 2006
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.17061759531497955,
      "learning_rate": 0.00014058065284183358,
      "loss": 1.0343,
      "step": 2007
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13605917990207672,
      "learning_rate": 0.00014052368628827058,
      "loss": 1.1489,
      "step": 2008
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13361580669879913,
      "learning_rate": 0.00014046670399577477,
      "loss": 1.103,
      "step": 2009
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1289708912372589,
      "learning_rate": 0.0001404097059864774,
      "loss": 1.1498,
      "step": 2010
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1475292593240738,
      "learning_rate": 0.00014035269228251587,
      "loss": 1.1306,
      "step": 2011
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.15678825974464417,
      "learning_rate": 0.00014029566290603367,
      "loss": 1.1086,
      "step": 2012
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.12211201339960098,
      "learning_rate": 0.00014023861787918032,
      "loss": 1.1707,
      "step": 2013
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13817457854747772,
      "learning_rate": 0.00014018155722411144,
      "loss": 1.1272,
      "step": 2014
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.11501845717430115,
      "learning_rate": 0.00014012448096298874,
      "loss": 1.1536,
      "step": 2015
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.10257050395011902,
      "learning_rate": 0.00014006738911798,
      "loss": 1.1523,
      "step": 2016
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.16586090624332428,
      "learning_rate": 0.000140010281711259,
      "loss": 1.1901,
      "step": 2017
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.15896005928516388,
      "learning_rate": 0.00013995315876500565,
      "loss": 1.2034,
      "step": 2018
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.14290226995944977,
      "learning_rate": 0.0001398960203014058,
      "loss": 1.1081,
      "step": 2019
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13846911489963531,
      "learning_rate": 0.0001398388663426514,
      "loss": 1.126,
      "step": 2020
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.10678835213184357,
      "learning_rate": 0.00013978169691094037,
      "loss": 1.2035,
      "step": 2021
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.14401929080486298,
      "learning_rate": 0.00013972451202847664,
      "loss": 1.2285,
      "step": 2022
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.18416672945022583,
      "learning_rate": 0.00013966731171747023,
      "loss": 1.1578,
      "step": 2023
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13061054050922394,
      "learning_rate": 0.000139610096000137,
      "loss": 1.1663,
      "step": 2024
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.11208663880825043,
      "learning_rate": 0.00013955286489869895,
      "loss": 1.2466,
      "step": 2025
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.12579205632209778,
      "learning_rate": 0.0001394956184353839,
      "loss": 1.1543,
      "step": 2026
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1083642989397049,
      "learning_rate": 0.00013943835663242577,
      "loss": 1.2331,
      "step": 2027
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.11667551100254059,
      "learning_rate": 0.00013938107951206439,
      "loss": 1.1246,
      "step": 2028
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.11943831294775009,
      "learning_rate": 0.00013932378709654547,
      "loss": 1.1286,
      "step": 2029
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.10436496138572693,
      "learning_rate": 0.00013926647940812082,
      "loss": 1.1338,
      "step": 2030
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1367201805114746,
      "learning_rate": 0.000139209156469048,
      "loss": 1.2115,
      "step": 2031
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.11488649994134903,
      "learning_rate": 0.00013915181830159061,
      "loss": 1.0954,
      "step": 2032
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.14618755877017975,
      "learning_rate": 0.00013909446492801818,
      "loss": 1.1267,
      "step": 2033
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1350533366203308,
      "learning_rate": 0.00013903709637060605,
      "loss": 1.0539,
      "step": 2034
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.130876362323761,
      "learning_rate": 0.00013897971265163544,
      "loss": 1.1047,
      "step": 2035
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.11723452806472778,
      "learning_rate": 0.00013892231379339368,
      "loss": 1.1225,
      "step": 2036
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.12529751658439636,
      "learning_rate": 0.00013886489981817374,
      "loss": 1.2522,
      "step": 2037
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.11815999448299408,
      "learning_rate": 0.00013880747074827455,
      "loss": 1.1661,
      "step": 2038
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13797582685947418,
      "learning_rate": 0.00013875002660600086,
      "loss": 1.1694,
      "step": 2039
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.15323777496814728,
      "learning_rate": 0.00013869256741366338,
      "loss": 1.1582,
      "step": 2040
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.12146853655576706,
      "learning_rate": 0.00013863509319357858,
      "loss": 1.2084,
      "step": 2041
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.09680850058794022,
      "learning_rate": 0.00013857760396806875,
      "loss": 1.1445,
      "step": 2042
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.10953642427921295,
      "learning_rate": 0.00013852009975946208,
      "loss": 1.1433,
      "step": 2043
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.12124217301607132,
      "learning_rate": 0.00013846258059009253,
      "loss": 1.1881,
      "step": 2044
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.16952645778656006,
      "learning_rate": 0.0001384050464822999,
      "loss": 1.1867,
      "step": 2045
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.17906242609024048,
      "learning_rate": 0.0001383474974584297,
      "loss": 1.1476,
      "step": 2046
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.17906242609024048,
      "learning_rate": 0.0001383474974584297,
      "loss": 1.0746,
      "step": 2047
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13557608425617218,
      "learning_rate": 0.0001382899335408334,
      "loss": 1.0394,
      "step": 2048
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13003142178058624,
      "learning_rate": 0.00013823235475186814,
      "loss": 1.1182,
      "step": 2049
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.10389148443937302,
      "learning_rate": 0.00013817476111389684,
      "loss": 1.0583,
      "step": 2050
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.16500842571258545,
      "learning_rate": 0.00013811715264928825,
      "loss": 1.2139,
      "step": 2051
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.13634410500526428,
      "learning_rate": 0.00013805952938041673,
      "loss": 1.0796,
      "step": 2052
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.15014752745628357,
      "learning_rate": 0.00013800189132966257,
      "loss": 1.149,
      "step": 2053
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.12004932761192322,
      "learning_rate": 0.00013794423851941175,
      "loss": 1.2184,
      "step": 2054
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.11619910597801208,
      "learning_rate": 0.00013788657097205592,
      "loss": 1.1828,
      "step": 2055
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.15478438138961792,
      "learning_rate": 0.00013782888870999245,
      "loss": 1.1972,
      "step": 2056
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1211991086602211,
      "learning_rate": 0.0001377711917556245,
      "loss": 1.0913,
      "step": 2057
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1205533966422081,
      "learning_rate": 0.00013771348013136097,
      "loss": 1.16,
      "step": 2058
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.14126557111740112,
      "learning_rate": 0.00013765575385961626,
      "loss": 1.1534,
      "step": 2059
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.11261311918497086,
      "learning_rate": 0.0001375980129628107,
      "loss": 1.0641,
      "step": 2060
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1557186245918274,
      "learning_rate": 0.00013754025746337014,
      "loss": 1.0809,
      "step": 2061
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.11005322635173798,
      "learning_rate": 0.00013748248738372615,
      "loss": 1.1525,
      "step": 2062
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13805246353149414,
      "learning_rate": 0.00013742470274631598,
      "loss": 1.1385,
      "step": 2063
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.15767881274223328,
      "learning_rate": 0.00013736690357358253,
      "loss": 1.0985,
      "step": 2064
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1224205270409584,
      "learning_rate": 0.00013730908988797427,
      "loss": 1.1182,
      "step": 2065
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.14242276549339294,
      "learning_rate": 0.00013725126171194543,
      "loss": 1.0841,
      "step": 2066
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.15476511418819427,
      "learning_rate": 0.0001371934190679558,
      "loss": 1.157,
      "step": 2067
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1342262476682663,
      "learning_rate": 0.00013713556197847074,
      "loss": 1.1702,
      "step": 2068
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13114023208618164,
      "learning_rate": 0.00013707769046596136,
      "loss": 1.1332,
      "step": 2069
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13548867404460907,
      "learning_rate": 0.00013701980455290425,
      "loss": 1.0816,
      "step": 2070
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1259010285139084,
      "learning_rate": 0.00013696190426178162,
      "loss": 1.1831,
      "step": 2071
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.11700521409511566,
      "learning_rate": 0.00013690398961508127,
      "loss": 1.2379,
      "step": 2072
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1197250708937645,
      "learning_rate": 0.00013684606063529663,
      "loss": 1.1064,
      "step": 2073
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.14340505003929138,
      "learning_rate": 0.00013678811734492657,
      "loss": 1.1068,
      "step": 2074
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13067230582237244,
      "learning_rate": 0.00013673015976647568,
      "loss": 1.1208,
      "step": 2075
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.15675394237041473,
      "learning_rate": 0.000136672187922454,
      "loss": 1.2549,
      "step": 2076
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.14556171000003815,
      "learning_rate": 0.00013661420183537704,
      "loss": 1.1489,
      "step": 2077
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1323160082101822,
      "learning_rate": 0.00013655620152776604,
      "loss": 1.0917,
      "step": 2078
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13322418928146362,
      "learning_rate": 0.0001364981870221476,
      "loss": 1.1563,
      "step": 2079
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.14634771645069122,
      "learning_rate": 0.00013644015834105388,
      "loss": 1.1683,
      "step": 2080
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13327878713607788,
      "learning_rate": 0.00013638211550702256,
      "loss": 1.1225,
      "step": 2081
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13380080461502075,
      "learning_rate": 0.0001363240585425968,
      "loss": 1.1704,
      "step": 2082
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.12085270881652832,
      "learning_rate": 0.0001362659874703253,
      "loss": 1.0615,
      "step": 2083
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.08904567360877991,
      "learning_rate": 0.00013620790231276214,
      "loss": 1.1676,
      "step": 2084
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13217034935951233,
      "learning_rate": 0.00013614980309246692,
      "loss": 1.1546,
      "step": 2085
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13338513672351837,
      "learning_rate": 0.00013609168983200472,
      "loss": 1.1698,
      "step": 2086
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1574154496192932,
      "learning_rate": 0.00013603356255394614,
      "loss": 1.089,
      "step": 2087
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.12549245357513428,
      "learning_rate": 0.00013597542128086703,
      "loss": 1.1647,
      "step": 2088
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.11596287041902542,
      "learning_rate": 0.00013591726603534885,
      "loss": 1.1568,
      "step": 2089
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.09910865873098373,
      "learning_rate": 0.00013585909683997842,
      "loss": 1.1466,
      "step": 2090
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.11457400023937225,
      "learning_rate": 0.00013580091371734798,
      "loss": 1.1517,
      "step": 2091
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.13053573668003082,
      "learning_rate": 0.00013574271669005518,
      "loss": 1.128,
      "step": 2092
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1435869038105011,
      "learning_rate": 0.00013568450578070308,
      "loss": 1.129,
      "step": 2093
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.09109068661928177,
      "learning_rate": 0.00013562628101190013,
      "loss": 1.1852,
      "step": 2094
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.12373720109462738,
      "learning_rate": 0.00013556804240626017,
      "loss": 1.1531,
      "step": 2095
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.12492918968200684,
      "learning_rate": 0.0001355097899864024,
      "loss": 1.1063,
      "step": 2096
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1379762589931488,
      "learning_rate": 0.00013545152377495136,
      "loss": 1.1292,
      "step": 2097
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.10757533460855484,
      "learning_rate": 0.00013539324379453698,
      "loss": 1.2266,
      "step": 2098
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1765671819448471,
      "learning_rate": 0.00013533495006779454,
      "loss": 1.0948,
      "step": 2099
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.10907498002052307,
      "learning_rate": 0.00013527664261736471,
      "loss": 1.17,
      "step": 2100
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.14740823209285736,
      "learning_rate": 0.00013521832146589334,
      "loss": 1.1942,
      "step": 2101
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.142926424741745,
      "learning_rate": 0.00013515998663603173,
      "loss": 1.1851,
      "step": 2102
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1356813907623291,
      "learning_rate": 0.00013510163815043646,
      "loss": 1.2268,
      "step": 2103
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.1270919144153595,
      "learning_rate": 0.00013504327603176944,
      "loss": 1.0842,
      "step": 2104
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.12729932367801666,
      "learning_rate": 0.0001349849003026978,
      "loss": 1.1123,
      "step": 2105
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.11695495247840881,
      "learning_rate": 0.00013492651098589398,
      "loss": 1.1958,
      "step": 2106
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12562772631645203,
      "learning_rate": 0.0001348681081040358,
      "loss": 1.0627,
      "step": 2107
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1424725502729416,
      "learning_rate": 0.0001348096916798062,
      "loss": 1.1508,
      "step": 2108
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09645726531744003,
      "learning_rate": 0.00013475126173589344,
      "loss": 1.0391,
      "step": 2109
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12540273368358612,
      "learning_rate": 0.00013469281829499107,
      "loss": 1.1167,
      "step": 2110
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.11971380561590195,
      "learning_rate": 0.00013463436137979787,
      "loss": 1.1193,
      "step": 2111
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1322985291481018,
      "learning_rate": 0.00013457589101301774,
      "loss": 1.186,
      "step": 2112
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1450778692960739,
      "learning_rate": 0.00013451740721736002,
      "loss": 1.1399,
      "step": 2113
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.13415852189064026,
      "learning_rate": 0.00013445891001553906,
      "loss": 1.0579,
      "step": 2114
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1558161973953247,
      "learning_rate": 0.00013440039943027451,
      "loss": 1.1573,
      "step": 2115
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12697958946228027,
      "learning_rate": 0.00013434187548429124,
      "loss": 1.1382,
      "step": 2116
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1306530088186264,
      "learning_rate": 0.00013428333820031922,
      "loss": 1.151,
      "step": 2117
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09952905029058456,
      "learning_rate": 0.00013422478760109373,
      "loss": 1.1572,
      "step": 2118
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.13804303109645844,
      "learning_rate": 0.00013416622370935507,
      "loss": 1.1405,
      "step": 2119
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12581968307495117,
      "learning_rate": 0.00013410764654784885,
      "loss": 1.1321,
      "step": 2120
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.17283649742603302,
      "learning_rate": 0.00013404905613932574,
      "loss": 1.1396,
      "step": 2121
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.15690481662750244,
      "learning_rate": 0.00013399045250654152,
      "loss": 1.1243,
      "step": 2122
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12063764035701752,
      "learning_rate": 0.00013393183567225725,
      "loss": 1.0533,
      "step": 2123
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12951348721981049,
      "learning_rate": 0.000133873205659239,
      "loss": 1.1651,
      "step": 2124
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12327460944652557,
      "learning_rate": 0.000133814562490258,
      "loss": 1.1336,
      "step": 2125
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.17766256630420685,
      "learning_rate": 0.00013375590618809054,
      "loss": 1.144,
      "step": 2126
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.13081716001033783,
      "learning_rate": 0.00013369723677551814,
      "loss": 1.2097,
      "step": 2127
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.11287351697683334,
      "learning_rate": 0.00013363855427532724,
      "loss": 1.0515,
      "step": 2128
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1293027549982071,
      "learning_rate": 0.00013357985871030947,
      "loss": 1.068,
      "step": 2129
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.14160916209220886,
      "learning_rate": 0.00013352115010326155,
      "loss": 1.0933,
      "step": 2130
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12530888617038727,
      "learning_rate": 0.00013346242847698517,
      "loss": 1.0658,
      "step": 2131
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.16500763595104218,
      "learning_rate": 0.00013340369385428713,
      "loss": 1.1462,
      "step": 2132
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.14351238310337067,
      "learning_rate": 0.00013334494625797936,
      "loss": 1.1375,
      "step": 2133
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12578167021274567,
      "learning_rate": 0.00013328618571087868,
      "loss": 1.1203,
      "step": 2134
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12659971415996552,
      "learning_rate": 0.000133227412235807,
      "loss": 1.194,
      "step": 2135
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.14161108434200287,
      "learning_rate": 0.0001331686258555913,
      "loss": 1.1449,
      "step": 2136
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12412907183170319,
      "learning_rate": 0.00013310982659306352,
      "loss": 1.1775,
      "step": 2137
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.10377461463212967,
      "learning_rate": 0.00013305101447106063,
      "loss": 1.1895,
      "step": 2138
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.14955683052539825,
      "learning_rate": 0.00013299218951242457,
      "loss": 1.1294,
      "step": 2139
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.14926540851593018,
      "learning_rate": 0.00013293335174000226,
      "loss": 1.1383,
      "step": 2140
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.10798218101263046,
      "learning_rate": 0.0001328745011766456,
      "loss": 1.112,
      "step": 2141
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09920880198478699,
      "learning_rate": 0.00013281563784521154,
      "loss": 1.1001,
      "step": 2142
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.13106068968772888,
      "learning_rate": 0.00013275676176856185,
      "loss": 1.1823,
      "step": 2143
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.13585792481899261,
      "learning_rate": 0.00013269787296956334,
      "loss": 1.1172,
      "step": 2144
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12461868673563004,
      "learning_rate": 0.00013263897147108776,
      "loss": 1.0866,
      "step": 2145
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.10043303668498993,
      "learning_rate": 0.00013258005729601177,
      "loss": 1.0897,
      "step": 2146
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.16691963374614716,
      "learning_rate": 0.00013252113046721692,
      "loss": 1.1826,
      "step": 2147
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1290479153394699,
      "learning_rate": 0.00013246219100758973,
      "loss": 1.1081,
      "step": 2148
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.13886041939258575,
      "learning_rate": 0.00013240323894002167,
      "loss": 1.1603,
      "step": 2149
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.09867934882640839,
      "learning_rate": 0.00013234427428740895,
      "loss": 1.1452,
      "step": 2150
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12843534350395203,
      "learning_rate": 0.00013228529707265277,
      "loss": 1.1424,
      "step": 2151
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.14377029240131378,
      "learning_rate": 0.00013222630731865928,
      "loss": 1.0255,
      "step": 2152
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.10712307691574097,
      "learning_rate": 0.00013216730504833935,
      "loss": 1.1657,
      "step": 2153
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.13084207475185394,
      "learning_rate": 0.00013210829028460883,
      "loss": 1.2215,
      "step": 2154
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.11641918122768402,
      "learning_rate": 0.00013204926305038832,
      "loss": 1.1793,
      "step": 2155
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.15262733399868011,
      "learning_rate": 0.00013199022336860333,
      "loss": 1.1806,
      "step": 2156
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.12558403611183167,
      "learning_rate": 0.00013193117126218424,
      "loss": 1.1236,
      "step": 2157
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.12102440744638443,
      "learning_rate": 0.00013187210675406617,
      "loss": 1.161,
      "step": 2158
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.12102440744638443,
      "learning_rate": 0.00013187210675406617,
      "loss": 1.1649,
      "step": 2159
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.12004001438617706,
      "learning_rate": 0.0001318130298671891,
      "loss": 1.188,
      "step": 2160
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.106879822909832,
      "learning_rate": 0.00013175394062449778,
      "loss": 1.1168,
      "step": 2161
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.13610805571079254,
      "learning_rate": 0.00013169483904894183,
      "loss": 1.0985,
      "step": 2162
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.10991246998310089,
      "learning_rate": 0.00013163572516347564,
      "loss": 1.1716,
      "step": 2163
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.10225801914930344,
      "learning_rate": 0.00013157659899105833,
      "loss": 1.0841,
      "step": 2164
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1138380914926529,
      "learning_rate": 0.0001315174605546538,
      "loss": 1.1449,
      "step": 2165
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14452305436134338,
      "learning_rate": 0.0001314583098772308,
      "loss": 1.1662,
      "step": 2166
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14590702950954437,
      "learning_rate": 0.00013139914698176273,
      "loss": 1.0948,
      "step": 2167
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.12272660434246063,
      "learning_rate": 0.00013133997189122777,
      "loss": 1.141,
      "step": 2168
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14340415596961975,
      "learning_rate": 0.00013128078462860886,
      "loss": 1.0657,
      "step": 2169
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14530819654464722,
      "learning_rate": 0.00013122158521689369,
      "loss": 1.0948,
      "step": 2170
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11325385421514511,
      "learning_rate": 0.00013116237367907454,
      "loss": 1.1415,
      "step": 2171
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1246979609131813,
      "learning_rate": 0.00013110315003814855,
      "loss": 1.1687,
      "step": 2172
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11029254645109177,
      "learning_rate": 0.00013104391431711747,
      "loss": 1.0991,
      "step": 2173
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.12244560569524765,
      "learning_rate": 0.0001309846665389878,
      "loss": 1.1817,
      "step": 2174
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11899139732122421,
      "learning_rate": 0.0001309254067267707,
      "loss": 1.1045,
      "step": 2175
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11526009440422058,
      "learning_rate": 0.00013086613490348198,
      "loss": 1.1145,
      "step": 2176
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14241886138916016,
      "learning_rate": 0.00013080685109214207,
      "loss": 1.1739,
      "step": 2177
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1328553855419159,
      "learning_rate": 0.00013074755531577626,
      "loss": 1.1141,
      "step": 2178
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.12329132109880447,
      "learning_rate": 0.00013068824759741426,
      "loss": 1.1497,
      "step": 2179
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.12720270454883575,
      "learning_rate": 0.0001306289279600905,
      "loss": 1.1649,
      "step": 2180
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1274811327457428,
      "learning_rate": 0.00013056959642684403,
      "loss": 1.0811,
      "step": 2181
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14049141108989716,
      "learning_rate": 0.0001305102530207186,
      "loss": 1.1752,
      "step": 2182
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11085979640483856,
      "learning_rate": 0.00013045089776476246,
      "loss": 1.1187,
      "step": 2183
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09479273110628128,
      "learning_rate": 0.0001303915306820285,
      "loss": 1.1949,
      "step": 2184
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14263056218624115,
      "learning_rate": 0.00013033215179557422,
      "loss": 1.0789,
      "step": 2185
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1387367695569992,
      "learning_rate": 0.0001302727611284617,
      "loss": 1.121,
      "step": 2186
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.13334740698337555,
      "learning_rate": 0.00013021335870375765,
      "loss": 1.0785,
      "step": 2187
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11391913890838623,
      "learning_rate": 0.00013015394454453315,
      "loss": 1.1841,
      "step": 2188
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11711546033620834,
      "learning_rate": 0.00013009451867386412,
      "loss": 1.2068,
      "step": 2189
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14356093108654022,
      "learning_rate": 0.00013003508111483075,
      "loss": 1.2121,
      "step": 2190
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09971053153276443,
      "learning_rate": 0.000129975631890518,
      "loss": 1.0819,
      "step": 2191
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11135859042406082,
      "learning_rate": 0.00012991617102401523,
      "loss": 1.1761,
      "step": 2192
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11385694891214371,
      "learning_rate": 0.00012985669853841634,
      "loss": 1.097,
      "step": 2193
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14653077721595764,
      "learning_rate": 0.0001297972144568198,
      "loss": 1.1197,
      "step": 2194
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.12157117575407028,
      "learning_rate": 0.00012973771880232853,
      "loss": 1.1324,
      "step": 2195
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1121988520026207,
      "learning_rate": 0.00012967821159804994,
      "loss": 1.1217,
      "step": 2196
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14949339628219604,
      "learning_rate": 0.00012961869286709594,
      "loss": 1.1077,
      "step": 2197
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.12020736187696457,
      "learning_rate": 0.0001295591626325829,
      "loss": 1.1563,
      "step": 2198
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11796382069587708,
      "learning_rate": 0.00012949962091763175,
      "loss": 1.112,
      "step": 2199
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11595911532640457,
      "learning_rate": 0.00012944006774536773,
      "loss": 1.1397,
      "step": 2200
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1418847143650055,
      "learning_rate": 0.00012938050313892062,
      "loss": 1.1942,
      "step": 2201
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.14700573682785034,
      "learning_rate": 0.00012932092712142468,
      "loss": 1.0984,
      "step": 2202
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.13308532536029816,
      "learning_rate": 0.00012926133971601848,
      "loss": 1.0446,
      "step": 2203
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09626142680644989,
      "learning_rate": 0.00012920174094584514,
      "loss": 1.0643,
      "step": 2204
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.16173294186592102,
      "learning_rate": 0.00012914213083405212,
      "loss": 1.0831,
      "step": 2205
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11041837930679321,
      "learning_rate": 0.00012908250940379124,
      "loss": 1.208,
      "step": 2206
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.09958840906620026,
      "learning_rate": 0.00012902287667821883,
      "loss": 1.0643,
      "step": 2207
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.13174206018447876,
      "learning_rate": 0.0001289632326804956,
      "loss": 1.1686,
      "step": 2208
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.11643630266189575,
      "learning_rate": 0.0001289035774337865,
      "loss": 1.0771,
      "step": 2209
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.20702525973320007,
      "learning_rate": 0.00012884391096126097,
      "loss": 1.0614,
      "step": 2210
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1372224986553192,
      "learning_rate": 0.00012878423328609281,
      "loss": 1.1303,
      "step": 2211
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.14591696858406067,
      "learning_rate": 0.00012872454443146013,
      "loss": 1.18,
      "step": 2212
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1293402910232544,
      "learning_rate": 0.00012866484442054538,
      "loss": 1.1295,
      "step": 2213
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1260233074426651,
      "learning_rate": 0.00012860513327653535,
      "loss": 1.1487,
      "step": 2214
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.09418804198503494,
      "learning_rate": 0.0001285454110226212,
      "loss": 1.1534,
      "step": 2215
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.12056389451026917,
      "learning_rate": 0.00012848567768199831,
      "loss": 1.1323,
      "step": 2216
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.18300761282444,
      "learning_rate": 0.00012842593327786646,
      "loss": 1.1626,
      "step": 2217
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.12067338824272156,
      "learning_rate": 0.0001283661778334297,
      "loss": 1.1444,
      "step": 2218
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.13589535653591156,
      "learning_rate": 0.00012830641137189628,
      "loss": 1.1451,
      "step": 2219
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1292678415775299,
      "learning_rate": 0.0001282466339164789,
      "loss": 1.138,
      "step": 2220
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1029343530535698,
      "learning_rate": 0.00012818684549039438,
      "loss": 1.1912,
      "step": 2221
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.10724412649869919,
      "learning_rate": 0.00012812704611686387,
      "loss": 0.9856,
      "step": 2222
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.15179774165153503,
      "learning_rate": 0.00012806723581911274,
      "loss": 1.1634,
      "step": 2223
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1391134411096573,
      "learning_rate": 0.00012800741462037064,
      "loss": 1.1953,
      "step": 2224
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1632150560617447,
      "learning_rate": 0.00012794758254387146,
      "loss": 1.1822,
      "step": 2225
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.125529944896698,
      "learning_rate": 0.00012788773961285323,
      "loss": 1.1126,
      "step": 2226
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.09500427544116974,
      "learning_rate": 0.00012782788585055828,
      "loss": 1.0483,
      "step": 2227
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.11065895855426788,
      "learning_rate": 0.00012776802128023316,
      "loss": 1.1329,
      "step": 2228
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.15528084337711334,
      "learning_rate": 0.0001277081459251285,
      "loss": 1.0656,
      "step": 2229
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.09463316202163696,
      "learning_rate": 0.0001276482598084993,
      "loss": 1.115,
      "step": 2230
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.12484826892614365,
      "learning_rate": 0.00012758836295360455,
      "loss": 1.0524,
      "step": 2231
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.13153918087482452,
      "learning_rate": 0.0001275284553837075,
      "loss": 1.1742,
      "step": 2232
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.14264151453971863,
      "learning_rate": 0.00012746853712207568,
      "loss": 1.2128,
      "step": 2233
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.10724753886461258,
      "learning_rate": 0.00012740860819198047,
      "loss": 1.0741,
      "step": 2234
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.13433967530727386,
      "learning_rate": 0.0001273486686166977,
      "loss": 1.1258,
      "step": 2235
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.15100201964378357,
      "learning_rate": 0.00012728871841950718,
      "loss": 1.1669,
      "step": 2236
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1364363580942154,
      "learning_rate": 0.00012722875762369288,
      "loss": 1.1176,
      "step": 2237
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1532536894083023,
      "learning_rate": 0.00012716878625254288,
      "loss": 1.0781,
      "step": 2238
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.13240830600261688,
      "learning_rate": 0.00012710880432934932,
      "loss": 1.232,
      "step": 2239
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.11691036075353622,
      "learning_rate": 0.0001270488118774086,
      "loss": 1.215,
      "step": 2240
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.12774819135665894,
      "learning_rate": 0.000126988808920021,
      "loss": 1.0581,
      "step": 2241
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.11066650599241257,
      "learning_rate": 0.000126928795480491,
      "loss": 1.1774,
      "step": 2242
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1210908368229866,
      "learning_rate": 0.00012686877158212713,
      "loss": 1.1206,
      "step": 2243
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.14309009909629822,
      "learning_rate": 0.00012680873724824198,
      "loss": 1.1064,
      "step": 2244
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.11371345072984695,
      "learning_rate": 0.00012674869250215223,
      "loss": 1.1804,
      "step": 2245
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.12694978713989258,
      "learning_rate": 0.00012668863736717856,
      "loss": 1.2025,
      "step": 2246
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.13588601350784302,
      "learning_rate": 0.00012662857186664559,
      "loss": 1.2035,
      "step": 2247
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.14391303062438965,
      "learning_rate": 0.0001265684960238822,
      "loss": 1.1508,
      "step": 2248
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.11524519324302673,
      "learning_rate": 0.0001265084098622211,
      "loss": 1.142,
      "step": 2249
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.10885654389858246,
      "learning_rate": 0.00012644831340499906,
      "loss": 1.1237,
      "step": 2250
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1429527997970581,
      "learning_rate": 0.00012638820667555683,
      "loss": 1.2119,
      "step": 2251
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.11710992455482483,
      "learning_rate": 0.00012632808969723926,
      "loss": 1.1208,
      "step": 2252
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.142898291349411,
      "learning_rate": 0.000126267962493395,
      "loss": 1.1739,
      "step": 2253
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1079142838716507,
      "learning_rate": 0.0001262078250873768,
      "loss": 1.1226,
      "step": 2254
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.13841184973716736,
      "learning_rate": 0.00012614767750254128,
      "loss": 1.166,
      "step": 2255
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.15410363674163818,
      "learning_rate": 0.00012608751976224915,
      "loss": 1.222,
      "step": 2256
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.11386985331773758,
      "learning_rate": 0.00012602735188986498,
      "loss": 1.1314,
      "step": 2257
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.12313192337751389,
      "learning_rate": 0.0001259671739087572,
      "loss": 1.1185,
      "step": 2258
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.1414235383272171,
      "learning_rate": 0.00012590698584229835,
      "loss": 1.1236,
      "step": 2259
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.16902711987495422,
      "learning_rate": 0.00012584678771386466,
      "loss": 1.1651,
      "step": 2260
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.15091834962368011,
      "learning_rate": 0.0001257865795468365,
      "loss": 1.1192,
      "step": 2261
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1767662614583969,
      "learning_rate": 0.000125726361364598,
      "loss": 1.0285,
      "step": 2262
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1299576312303543,
      "learning_rate": 0.00012566613319053713,
      "loss": 1.0543,
      "step": 2263
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.11902175843715668,
      "learning_rate": 0.00012560589504804592,
      "loss": 1.1344,
      "step": 2264
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.14813445508480072,
      "learning_rate": 0.00012554564696052011,
      "loss": 1.1941,
      "step": 2265
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12827658653259277,
      "learning_rate": 0.00012548538895135942,
      "loss": 1.1088,
      "step": 2266
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12653204798698425,
      "learning_rate": 0.00012542512104396728,
      "loss": 1.1931,
      "step": 2267
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.13994964957237244,
      "learning_rate": 0.00012536484326175113,
      "loss": 1.1505,
      "step": 2268
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.10577087849378586,
      "learning_rate": 0.00012530455562812214,
      "loss": 1.0502,
      "step": 2269
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.144480362534523,
      "learning_rate": 0.0001252442581664953,
      "loss": 1.0861,
      "step": 2270
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.16437703371047974,
      "learning_rate": 0.0001251839509002895,
      "loss": 1.1901,
      "step": 2271
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.14927423000335693,
      "learning_rate": 0.00012512363385292738,
      "loss": 1.1785,
      "step": 2272
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.23060639202594757,
      "learning_rate": 0.00012506330704783534,
      "loss": 1.1649,
      "step": 2273
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1362883746623993,
      "learning_rate": 0.00012500297050844366,
      "loss": 1.1757,
      "step": 2274
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1575712263584137,
      "learning_rate": 0.00012494262425818636,
      "loss": 1.1802,
      "step": 2275
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.11796757578849792,
      "learning_rate": 0.00012488226832050117,
      "loss": 1.0124,
      "step": 2276
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12970848381519318,
      "learning_rate": 0.00012482190271882973,
      "loss": 1.0317,
      "step": 2277
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.17965322732925415,
      "learning_rate": 0.00012476152747661727,
      "loss": 1.1359,
      "step": 2278
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.13923786580562592,
      "learning_rate": 0.00012470114261731288,
      "loss": 1.1938,
      "step": 2279
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.13842135667800903,
      "learning_rate": 0.0001246407481643693,
      "loss": 1.1511,
      "step": 2280
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1188579648733139,
      "learning_rate": 0.0001245803441412431,
      "loss": 1.1859,
      "step": 2281
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.13101018965244293,
      "learning_rate": 0.00012451993057139443,
      "loss": 1.2022,
      "step": 2282
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.13665038347244263,
      "learning_rate": 0.0001244595074782873,
      "loss": 1.1545,
      "step": 2283
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.14524051547050476,
      "learning_rate": 0.00012439907488538934,
      "loss": 1.1138,
      "step": 2284
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1273861527442932,
      "learning_rate": 0.0001243386328161718,
      "loss": 1.1528,
      "step": 2285
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12134937942028046,
      "learning_rate": 0.00012427818129410976,
      "loss": 1.1308,
      "step": 2286
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.10555349290370941,
      "learning_rate": 0.00012421772034268186,
      "loss": 1.212,
      "step": 2287
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.11301744729280472,
      "learning_rate": 0.00012415724998537041,
      "loss": 1.1763,
      "step": 2288
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12921105325222015,
      "learning_rate": 0.00012409677024566144,
      "loss": 1.1364,
      "step": 2289
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12318039685487747,
      "learning_rate": 0.00012403628114704459,
      "loss": 1.1168,
      "step": 2290
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1387970745563507,
      "learning_rate": 0.0001239757827130131,
      "loss": 1.2387,
      "step": 2291
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1825876086950302,
      "learning_rate": 0.00012391527496706387,
      "loss": 1.0953,
      "step": 2292
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.13448865711688995,
      "learning_rate": 0.00012385475793269744,
      "loss": 1.1193,
      "step": 2293
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.13687287271022797,
      "learning_rate": 0.0001237942316334179,
      "loss": 1.1392,
      "step": 2294
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.11378175765275955,
      "learning_rate": 0.00012373369609273296,
      "loss": 1.1549,
      "step": 2295
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12041927874088287,
      "learning_rate": 0.00012367315133415396,
      "loss": 1.0315,
      "step": 2296
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.10859940946102142,
      "learning_rate": 0.00012361259738119575,
      "loss": 1.0509,
      "step": 2297
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.14032043516635895,
      "learning_rate": 0.00012355203425737683,
      "loss": 1.1694,
      "step": 2298
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12932714819908142,
      "learning_rate": 0.00012349146198621918,
      "loss": 1.1685,
      "step": 2299
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12541405856609344,
      "learning_rate": 0.0001234308805912484,
      "loss": 1.224,
      "step": 2300
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1623315066099167,
      "learning_rate": 0.00012337029009599357,
      "loss": 1.1057,
      "step": 2301
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.10571138560771942,
      "learning_rate": 0.00012330969052398735,
      "loss": 1.0764,
      "step": 2302
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12863528728485107,
      "learning_rate": 0.00012324908189876595,
      "loss": 1.1629,
      "step": 2303
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.14393901824951172,
      "learning_rate": 0.00012318846424386908,
      "loss": 1.1188,
      "step": 2304
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1392608880996704,
      "learning_rate": 0.0001231278375828398,
      "loss": 1.1488,
      "step": 2305
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.14626729488372803,
      "learning_rate": 0.00012306720193922498,
      "loss": 1.1776,
      "step": 2306
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12175363302230835,
      "learning_rate": 0.00012300655733657473,
      "loss": 1.1559,
      "step": 2307
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1290426403284073,
      "learning_rate": 0.0001229459037984427,
      "loss": 1.118,
      "step": 2308
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.13575154542922974,
      "learning_rate": 0.000122885241348386,
      "loss": 1.0971,
      "step": 2309
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.12853455543518066,
      "learning_rate": 0.0001228245700099653,
      "loss": 1.1476,
      "step": 2310
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.14863377809524536,
      "learning_rate": 0.00012276388980674466,
      "loss": 1.163,
      "step": 2311
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.11735289543867111,
      "learning_rate": 0.0001227032007622915,
      "loss": 1.0958,
      "step": 2312
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12968474626541138,
      "learning_rate": 0.00012264250290017673,
      "loss": 1.1761,
      "step": 2313
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.14217860996723175,
      "learning_rate": 0.00012258179624397478,
      "loss": 1.0777,
      "step": 2314
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.16156260669231415,
      "learning_rate": 0.00012252108081726336,
      "loss": 1.1945,
      "step": 2315
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.17025546729564667,
      "learning_rate": 0.0001224603566436237,
      "loss": 1.0895,
      "step": 2316
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.145872563123703,
      "learning_rate": 0.00012239962374664028,
      "loss": 1.1623,
      "step": 2317
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1374092698097229,
      "learning_rate": 0.00012233888214990113,
      "loss": 1.1431,
      "step": 2318
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1561015248298645,
      "learning_rate": 0.00012227813187699757,
      "loss": 1.1129,
      "step": 2319
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1432427018880844,
      "learning_rate": 0.00012221737295152428,
      "loss": 1.2035,
      "step": 2320
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.14017903804779053,
      "learning_rate": 0.00012215660539707936,
      "loss": 1.1305,
      "step": 2321
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.158723846077919,
      "learning_rate": 0.00012209582923726424,
      "loss": 1.1429,
      "step": 2322
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.13946552574634552,
      "learning_rate": 0.00012203504449568362,
      "loss": 1.1302,
      "step": 2323
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.11974669992923737,
      "learning_rate": 0.00012197425119594563,
      "loss": 1.1401,
      "step": 2324
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.15260237455368042,
      "learning_rate": 0.00012191344936166169,
      "loss": 1.1155,
      "step": 2325
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.13295167684555054,
      "learning_rate": 0.00012185263901644653,
      "loss": 1.11,
      "step": 2326
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.13186781108379364,
      "learning_rate": 0.00012179182018391819,
      "loss": 1.141,
      "step": 2327
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.09180246293544769,
      "learning_rate": 0.00012173099288769798,
      "loss": 1.1106,
      "step": 2328
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1300136148929596,
      "learning_rate": 0.00012167015715141057,
      "loss": 1.0609,
      "step": 2329
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.16662122309207916,
      "learning_rate": 0.00012160931299868378,
      "loss": 1.1026,
      "step": 2330
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1189848855137825,
      "learning_rate": 0.00012154846045314884,
      "loss": 1.0895,
      "step": 2331
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1450086236000061,
      "learning_rate": 0.0001214875995384402,
      "loss": 1.1725,
      "step": 2332
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12755687534809113,
      "learning_rate": 0.00012142673027819549,
      "loss": 1.2312,
      "step": 2333
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.16386958956718445,
      "learning_rate": 0.00012136585269605558,
      "loss": 1.1289,
      "step": 2334
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12032952159643173,
      "learning_rate": 0.00012130496681566475,
      "loss": 1.078,
      "step": 2335
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.10992022603750229,
      "learning_rate": 0.00012124407266067028,
      "loss": 1.0917,
      "step": 2336
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.15815876424312592,
      "learning_rate": 0.00012118317025472279,
      "loss": 1.1308,
      "step": 2337
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12674210965633392,
      "learning_rate": 0.00012112225962147605,
      "loss": 1.1217,
      "step": 2338
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.14242546260356903,
      "learning_rate": 0.00012106134078458709,
      "loss": 1.1273,
      "step": 2339
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.15238024294376373,
      "learning_rate": 0.00012100041376771605,
      "loss": 1.1162,
      "step": 2340
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12611456215381622,
      "learning_rate": 0.0001209394785945263,
      "loss": 1.0827,
      "step": 2341
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.10943541675806046,
      "learning_rate": 0.00012087853528868431,
      "loss": 1.1851,
      "step": 2342
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.11915472149848938,
      "learning_rate": 0.00012081758387385982,
      "loss": 1.1634,
      "step": 2343
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12836043536663055,
      "learning_rate": 0.00012075662437372566,
      "loss": 1.0846,
      "step": 2344
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.15742133557796478,
      "learning_rate": 0.00012069565681195776,
      "loss": 1.1408,
      "step": 2345
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1460723578929901,
      "learning_rate": 0.0001206346812122352,
      "loss": 1.0421,
      "step": 2346
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.13024526834487915,
      "learning_rate": 0.00012057369759824024,
      "loss": 1.2035,
      "step": 2347
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12536542117595673,
      "learning_rate": 0.00012051270599365824,
      "loss": 1.1417,
      "step": 2348
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1712067574262619,
      "learning_rate": 0.00012045170642217755,
      "loss": 1.1516,
      "step": 2349
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.11725248396396637,
      "learning_rate": 0.00012039069890748978,
      "loss": 1.1318,
      "step": 2350
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.11683332175016403,
      "learning_rate": 0.00012032968347328953,
      "loss": 1.1454,
      "step": 2351
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.13688313961029053,
      "learning_rate": 0.00012026866014327444,
      "loss": 1.132,
      "step": 2352
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1343306303024292,
      "learning_rate": 0.00012020762894114535,
      "loss": 1.1226,
      "step": 2353
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.09963720291852951,
      "learning_rate": 0.00012014658989060599,
      "loss": 1.1255,
      "step": 2354
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12522055208683014,
      "learning_rate": 0.00012008554301536328,
      "loss": 1.1071,
      "step": 2355
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.13088448345661163,
      "learning_rate": 0.00012002448833912711,
      "loss": 1.1691,
      "step": 2356
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.09780855476856232,
      "learning_rate": 0.00011996342588561043,
      "loss": 1.1517,
      "step": 2357
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1466035544872284,
      "learning_rate": 0.00011990235567852917,
      "loss": 1.2488,
      "step": 2358
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.11980926990509033,
      "learning_rate": 0.00011984127774160225,
      "loss": 1.1273,
      "step": 2359
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.10129745304584503,
      "learning_rate": 0.00011978019209855174,
      "loss": 1.0965,
      "step": 2360
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1501641422510147,
      "learning_rate": 0.00011971909877310253,
      "loss": 1.1542,
      "step": 2361
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1543593406677246,
      "learning_rate": 0.00011965799778898256,
      "loss": 1.1461,
      "step": 2362
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12287788838148117,
      "learning_rate": 0.00011959688916992277,
      "loss": 1.1831,
      "step": 2363
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.12653201818466187,
      "learning_rate": 0.00011953577293965708,
      "loss": 1.1353,
      "step": 2364
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12817257642745972,
      "learning_rate": 0.00011947464912192227,
      "loss": 1.1091,
      "step": 2365
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1615428328514099,
      "learning_rate": 0.00011941351774045815,
      "loss": 1.1006,
      "step": 2366
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.13168393075466156,
      "learning_rate": 0.00011935237881900743,
      "loss": 1.2271,
      "step": 2367
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1557355523109436,
      "learning_rate": 0.00011929123238131579,
      "loss": 1.1548,
      "step": 2368
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.15055347979068756,
      "learning_rate": 0.00011923007845113176,
      "loss": 1.1649,
      "step": 2369
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12924352288246155,
      "learning_rate": 0.00011916891705220687,
      "loss": 1.1246,
      "step": 2370
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12210793048143387,
      "learning_rate": 0.00011910774820829549,
      "loss": 1.1374,
      "step": 2371
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.09031463414430618,
      "learning_rate": 0.00011904657194315485,
      "loss": 1.0989,
      "step": 2372
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1490110605955124,
      "learning_rate": 0.00011898538828054517,
      "loss": 1.1733,
      "step": 2373
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.14006853103637695,
      "learning_rate": 0.00011892419724422946,
      "loss": 1.1734,
      "step": 2374
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.10786260664463043,
      "learning_rate": 0.00011886299885797356,
      "loss": 1.0886,
      "step": 2375
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.17631562054157257,
      "learning_rate": 0.00011880179314554629,
      "loss": 1.1067,
      "step": 2376
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.10919833928346634,
      "learning_rate": 0.00011874058013071923,
      "loss": 1.2483,
      "step": 2377
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.11786956340074539,
      "learning_rate": 0.00011867935983726677,
      "loss": 1.1603,
      "step": 2378
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12652313709259033,
      "learning_rate": 0.0001186181322889662,
      "loss": 1.2086,
      "step": 2379
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12094798684120178,
      "learning_rate": 0.00011855689750959757,
      "loss": 1.1508,
      "step": 2380
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.14241310954093933,
      "learning_rate": 0.0001184956555229438,
      "loss": 1.1342,
      "step": 2381
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.13484342396259308,
      "learning_rate": 0.00011843440635279055,
      "loss": 1.1898,
      "step": 2382
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.13709622621536255,
      "learning_rate": 0.00011837315002292628,
      "loss": 1.1356,
      "step": 2383
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.15195201337337494,
      "learning_rate": 0.00011831188655714225,
      "loss": 1.0696,
      "step": 2384
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.16301321983337402,
      "learning_rate": 0.0001182506159792325,
      "loss": 1.1124,
      "step": 2385
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12236084789037704,
      "learning_rate": 0.00011818933831299382,
      "loss": 1.1367,
      "step": 2386
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.15181849896907806,
      "learning_rate": 0.00011812805358222571,
      "loss": 1.1575,
      "step": 2387
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.14942988753318787,
      "learning_rate": 0.00011806676181073048,
      "loss": 1.1586,
      "step": 2388
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1113707646727562,
      "learning_rate": 0.00011800546302231316,
      "loss": 1.1,
      "step": 2389
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.11169029027223587,
      "learning_rate": 0.00011794415724078147,
      "loss": 1.1501,
      "step": 2390
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.11151497811079025,
      "learning_rate": 0.00011788284448994586,
      "loss": 1.1881,
      "step": 2391
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.11820952594280243,
      "learning_rate": 0.00011782152479361957,
      "loss": 1.156,
      "step": 2392
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.18733540177345276,
      "learning_rate": 0.00011776019817561834,
      "loss": 1.0242,
      "step": 2393
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1303071528673172,
      "learning_rate": 0.00011769886465976086,
      "loss": 1.0925,
      "step": 2394
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.16862043738365173,
      "learning_rate": 0.00011763752426986823,
      "loss": 1.2397,
      "step": 2395
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.14149954915046692,
      "learning_rate": 0.00011757617702976442,
      "loss": 1.1498,
      "step": 2396
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12710857391357422,
      "learning_rate": 0.000117514822963276,
      "loss": 1.2313,
      "step": 2397
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12578560411930084,
      "learning_rate": 0.00011745346209423216,
      "loss": 1.1046,
      "step": 2398
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12293757498264313,
      "learning_rate": 0.00011739209444646477,
      "loss": 1.1555,
      "step": 2399
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1208065003156662,
      "learning_rate": 0.00011733072004380827,
      "loss": 1.2073,
      "step": 2400
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12296905368566513,
      "learning_rate": 0.00011726933891009983,
      "loss": 1.0871,
      "step": 2401
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.11967169493436813,
      "learning_rate": 0.00011720795106917917,
      "loss": 1.1589,
      "step": 2402
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.09815314412117004,
      "learning_rate": 0.00011714655654488859,
      "loss": 1.1597,
      "step": 2403
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.11771894246339798,
      "learning_rate": 0.00011708515536107299,
      "loss": 1.1475,
      "step": 2404
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1402588039636612,
      "learning_rate": 0.00011702374754157997,
      "loss": 1.0821,
      "step": 2405
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.15301692485809326,
      "learning_rate": 0.00011696233311025957,
      "loss": 1.1782,
      "step": 2406
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.16373823583126068,
      "learning_rate": 0.00011690091209096441,
      "loss": 1.065,
      "step": 2407
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.12606607377529144,
      "learning_rate": 0.00011683948450754976,
      "loss": 1.1518,
      "step": 2408
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.17001375555992126,
      "learning_rate": 0.00011677805038387337,
      "loss": 1.13,
      "step": 2409
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.19809626042842865,
      "learning_rate": 0.00011671660974379555,
      "loss": 1.1894,
      "step": 2410
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.14150205254554749,
      "learning_rate": 0.00011665516261117912,
      "loss": 1.0823,
      "step": 2411
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.16068094968795776,
      "learning_rate": 0.00011659370900988947,
      "loss": 1.1292,
      "step": 2412
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1485852599143982,
      "learning_rate": 0.00011653224896379439,
      "loss": 1.0908,
      "step": 2413
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1500120311975479,
      "learning_rate": 0.0001164707824967644,
      "loss": 1.1719,
      "step": 2414
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1148294135928154,
      "learning_rate": 0.00011640930963267225,
      "loss": 1.1583,
      "step": 2415
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.10786499083042145,
      "learning_rate": 0.00011634783039539328,
      "loss": 1.1813,
      "step": 2416
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1278669536113739,
      "learning_rate": 0.0001162863448088054,
      "loss": 1.0783,
      "step": 2417
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.12975914776325226,
      "learning_rate": 0.00011622485289678886,
      "loss": 1.0531,
      "step": 2418
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1401025354862213,
      "learning_rate": 0.00011616335468322641,
      "loss": 1.0671,
      "step": 2419
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.13023021817207336,
      "learning_rate": 0.00011610185019200323,
      "loss": 1.1599,
      "step": 2420
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.14938633143901825,
      "learning_rate": 0.00011604033944700701,
      "loss": 1.1551,
      "step": 2421
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.13578148186206818,
      "learning_rate": 0.00011597882247212776,
      "loss": 1.0797,
      "step": 2422
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.14894872903823853,
      "learning_rate": 0.000115917299291258,
      "loss": 1.1137,
      "step": 2423
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.12851135432720184,
      "learning_rate": 0.00011585576992829261,
      "loss": 1.1305,
      "step": 2424
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.18900051712989807,
      "learning_rate": 0.00011579423440712886,
      "loss": 1.2207,
      "step": 2425
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1257353574037552,
      "learning_rate": 0.00011573269275166652,
      "loss": 1.2425,
      "step": 2426
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10859306156635284,
      "learning_rate": 0.00011567114498580758,
      "loss": 1.0847,
      "step": 2427
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1569720208644867,
      "learning_rate": 0.00011560959113345649,
      "loss": 1.1969,
      "step": 2428
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.12020371854305267,
      "learning_rate": 0.00011554803121852005,
      "loss": 1.0677,
      "step": 2429
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.14792566001415253,
      "learning_rate": 0.00011548646526490749,
      "loss": 1.1291,
      "step": 2430
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1536005437374115,
      "learning_rate": 0.00011542489329653024,
      "loss": 1.1343,
      "step": 2431
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.16265130043029785,
      "learning_rate": 0.00011536331533730221,
      "loss": 1.2212,
      "step": 2432
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1403048187494278,
      "learning_rate": 0.00011530173141113947,
      "loss": 1.1125,
      "step": 2433
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.11875393241643906,
      "learning_rate": 0.00011524014154196063,
      "loss": 1.0928,
      "step": 2434
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.12558819353580475,
      "learning_rate": 0.00011517854575368642,
      "loss": 1.1115,
      "step": 2435
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10983597487211227,
      "learning_rate": 0.00011511694407023994,
      "loss": 1.1805,
      "step": 2436
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.17416618764400482,
      "learning_rate": 0.00011505533651554654,
      "loss": 1.1491,
      "step": 2437
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.11214861273765564,
      "learning_rate": 0.00011499372311353398,
      "loss": 1.2476,
      "step": 2438
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1305060237646103,
      "learning_rate": 0.0001149321038881321,
      "loss": 1.2067,
      "step": 2439
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.12850059568881989,
      "learning_rate": 0.00011487047886327313,
      "loss": 1.1255,
      "step": 2440
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.12527866661548615,
      "learning_rate": 0.00011480884806289152,
      "loss": 1.1431,
      "step": 2441
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1570521742105484,
      "learning_rate": 0.00011474721151092396,
      "loss": 1.1419,
      "step": 2442
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.16989506781101227,
      "learning_rate": 0.00011468556923130942,
      "loss": 1.1218,
      "step": 2443
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.14044888317584991,
      "learning_rate": 0.000114623921247989,
      "loss": 1.0764,
      "step": 2444
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.12141896784305573,
      "learning_rate": 0.00011456226758490602,
      "loss": 1.0681,
      "step": 2445
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1303868442773819,
      "learning_rate": 0.00011450060826600617,
      "loss": 1.2304,
      "step": 2446
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1283612996339798,
      "learning_rate": 0.00011443894331523717,
      "loss": 1.2233,
      "step": 2447
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1395663172006607,
      "learning_rate": 0.00011437727275654892,
      "loss": 1.0897,
      "step": 2448
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10994945466518402,
      "learning_rate": 0.00011431559661389362,
      "loss": 1.0722,
      "step": 2449
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.15006354451179504,
      "learning_rate": 0.00011425391491122556,
      "loss": 1.2831,
      "step": 2450
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10055369883775711,
      "learning_rate": 0.00011419222767250119,
      "loss": 1.0993,
      "step": 2451
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1593678891658783,
      "learning_rate": 0.00011413053492167916,
      "loss": 1.2483,
      "step": 2452
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.13200093805789948,
      "learning_rate": 0.00011406883668272014,
      "loss": 1.2133,
      "step": 2453
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.15597222745418549,
      "learning_rate": 0.0001140071329795871,
      "loss": 1.1021,
      "step": 2454
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.143216073513031,
      "learning_rate": 0.000113945423836245,
      "loss": 1.0818,
      "step": 2455
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.11809000372886658,
      "learning_rate": 0.00011388370927666101,
      "loss": 1.2324,
      "step": 2456
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1715926080942154,
      "learning_rate": 0.00011382198932480428,
      "loss": 1.1206,
      "step": 2457
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.11396922171115875,
      "learning_rate": 0.00011376026400464617,
      "loss": 1.097,
      "step": 2458
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.128747820854187,
      "learning_rate": 0.00011369853334016009,
      "loss": 1.114,
      "step": 2459
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.14283747971057892,
      "learning_rate": 0.0001136367973553215,
      "loss": 1.0789,
      "step": 2460
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10895591229200363,
      "learning_rate": 0.00011357505607410797,
      "loss": 1.1976,
      "step": 2461
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10969946533441544,
      "learning_rate": 0.00011351330952049908,
      "loss": 1.1009,
      "step": 2462
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10337924212217331,
      "learning_rate": 0.00011345155771847646,
      "loss": 1.0737,
      "step": 2463
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.13042570650577545,
      "learning_rate": 0.00011338980069202387,
      "loss": 1.1187,
      "step": 2464
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.10978246480226517,
      "learning_rate": 0.00011332803846512697,
      "loss": 1.1046,
      "step": 2465
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.13342539966106415,
      "learning_rate": 0.00011326627106177347,
      "loss": 1.193,
      "step": 2466
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.12094882130622864,
      "learning_rate": 0.0001132044985059532,
      "loss": 1.1247,
      "step": 2467
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.11609942466020584,
      "learning_rate": 0.00011314272082165784,
      "loss": 1.216,
      "step": 2468
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.13008546829223633,
      "learning_rate": 0.00011308093803288118,
      "loss": 1.1037,
      "step": 2469
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.10596262663602829,
      "learning_rate": 0.00011301915016361891,
      "loss": 1.0303,
      "step": 2470
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12269974499940872,
      "learning_rate": 0.0001129573572378687,
      "loss": 1.1603,
      "step": 2471
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11354339867830276,
      "learning_rate": 0.00011289555927963031,
      "loss": 1.1123,
      "step": 2472
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.14776398241519928,
      "learning_rate": 0.00011283375631290527,
      "loss": 1.1175,
      "step": 2473
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.13175372779369354,
      "learning_rate": 0.00011277194836169714,
      "loss": 1.1121,
      "step": 2474
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11571395397186279,
      "learning_rate": 0.00011271013545001143,
      "loss": 1.1151,
      "step": 2475
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1289510428905487,
      "learning_rate": 0.00011264831760185562,
      "loss": 1.1284,
      "step": 2476
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.17575794458389282,
      "learning_rate": 0.00011258649484123896,
      "loss": 1.1194,
      "step": 2477
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11179293692111969,
      "learning_rate": 0.00011252466719217273,
      "loss": 1.1083,
      "step": 2478
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12331674247980118,
      "learning_rate": 0.00011246283467867012,
      "loss": 1.1064,
      "step": 2479
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12041114270687103,
      "learning_rate": 0.00011240099732474611,
      "loss": 1.168,
      "step": 2480
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1282188892364502,
      "learning_rate": 0.00011233915515441764,
      "loss": 1.1293,
      "step": 2481
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11938702315092087,
      "learning_rate": 0.00011227730819170348,
      "loss": 1.13,
      "step": 2482
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1325252801179886,
      "learning_rate": 0.0001122154564606243,
      "loss": 1.2361,
      "step": 2483
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.09854891151189804,
      "learning_rate": 0.0001121535999852026,
      "loss": 1.1632,
      "step": 2484
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1475270390510559,
      "learning_rate": 0.0001120917387894627,
      "loss": 1.2053,
      "step": 2485
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.14119137823581696,
      "learning_rate": 0.00011202987289743079,
      "loss": 1.0983,
      "step": 2486
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11794048547744751,
      "learning_rate": 0.00011196800233313487,
      "loss": 1.2526,
      "step": 2487
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.15964743494987488,
      "learning_rate": 0.00011190612712060476,
      "loss": 1.1411,
      "step": 2488
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.08902118355035782,
      "learning_rate": 0.00011184424728387204,
      "loss": 1.1168,
      "step": 2489
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11586875468492508,
      "learning_rate": 0.00011178236284697017,
      "loss": 1.1175,
      "step": 2490
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.09514512121677399,
      "learning_rate": 0.00011172047383393434,
      "loss": 1.1236,
      "step": 2491
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11379586160182953,
      "learning_rate": 0.00011165858026880152,
      "loss": 1.0101,
      "step": 2492
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1506500095129013,
      "learning_rate": 0.00011159668217561048,
      "loss": 1.1149,
      "step": 2493
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11903068423271179,
      "learning_rate": 0.0001115347795784017,
      "loss": 1.2376,
      "step": 2494
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12342535704374313,
      "learning_rate": 0.00011147287250121745,
      "loss": 1.1303,
      "step": 2495
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11313383281230927,
      "learning_rate": 0.00011141096096810173,
      "loss": 1.1226,
      "step": 2496
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12434618175029755,
      "learning_rate": 0.00011134904500310028,
      "loss": 1.1545,
      "step": 2497
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.18530333042144775,
      "learning_rate": 0.00011128712463026048,
      "loss": 1.3171,
      "step": 2498
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11000332981348038,
      "learning_rate": 0.00011122519987363155,
      "loss": 1.1429,
      "step": 2499
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.13680435717105865,
      "learning_rate": 0.00011116327075726435,
      "loss": 1.1553,
      "step": 2500
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.14944526553153992,
      "learning_rate": 0.00011110133730521142,
      "loss": 1.0945,
      "step": 2501
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.13274356722831726,
      "learning_rate": 0.000111039399541527,
      "loss": 1.1869,
      "step": 2502
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.14201316237449646,
      "learning_rate": 0.000110977457490267,
      "loss": 1.198,
      "step": 2503
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12138894945383072,
      "learning_rate": 0.000110915511175489,
      "loss": 1.1991,
      "step": 2504
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11839750409126282,
      "learning_rate": 0.00011085356062125225,
      "loss": 1.1528,
      "step": 2505
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.10953894257545471,
      "learning_rate": 0.00011079160585161759,
      "loss": 1.1171,
      "step": 2506
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.13790211081504822,
      "learning_rate": 0.00011072964689064759,
      "loss": 1.0998,
      "step": 2507
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11405950039625168,
      "learning_rate": 0.0001106676837624064,
      "loss": 1.2239,
      "step": 2508
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12191779166460037,
      "learning_rate": 0.00011060571649095972,
      "loss": 1.2428,
      "step": 2509
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1722533106803894,
      "learning_rate": 0.000110543745100375,
      "loss": 1.0442,
      "step": 2510
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.15299466252326965,
      "learning_rate": 0.00011048176961472114,
      "loss": 1.2081,
      "step": 2511
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.12876176834106445,
      "learning_rate": 0.00011041979005806875,
      "loss": 1.0767,
      "step": 2512
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.16778993606567383,
      "learning_rate": 0.00011035780645449001,
      "loss": 1.1323,
      "step": 2513
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.120756134390831,
      "learning_rate": 0.00011029581882805857,
      "loss": 1.0783,
      "step": 2514
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.1072695180773735,
      "learning_rate": 0.00011023382720284972,
      "loss": 1.0743,
      "step": 2515
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.13450108468532562,
      "learning_rate": 0.00011017183160294033,
      "loss": 1.0983,
      "step": 2516
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.13391394913196564,
      "learning_rate": 0.00011010983205240877,
      "loss": 1.1706,
      "step": 2517
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.14714689552783966,
      "learning_rate": 0.00011004782857533489,
      "loss": 1.0936,
      "step": 2518
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.13163542747497559,
      "learning_rate": 0.00010998582119580019,
      "loss": 1.1057,
      "step": 2519
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.14594964683055878,
      "learning_rate": 0.00010992380993788762,
      "loss": 1.0428,
      "step": 2520
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.11818338185548782,
      "learning_rate": 0.00010986179482568161,
      "loss": 1.1697,
      "step": 2521
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12941883504390717,
      "learning_rate": 0.00010979977588326815,
      "loss": 1.1212,
      "step": 2522
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.17774881422519684,
      "learning_rate": 0.00010973775313473464,
      "loss": 1.164,
      "step": 2523
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1487135887145996,
      "learning_rate": 0.00010967572660417,
      "loss": 1.123,
      "step": 2524
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.10504540801048279,
      "learning_rate": 0.00010961369631566467,
      "loss": 1.0675,
      "step": 2525
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.11167105287313461,
      "learning_rate": 0.00010955166229331048,
      "loss": 1.0382,
      "step": 2526
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12967988848686218,
      "learning_rate": 0.00010948962456120068,
      "loss": 1.1535,
      "step": 2527
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.11031866073608398,
      "learning_rate": 0.00010942758314343006,
      "loss": 1.0938,
      "step": 2528
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.13731195032596588,
      "learning_rate": 0.00010936553806409482,
      "loss": 1.1729,
      "step": 2529
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.15317000448703766,
      "learning_rate": 0.00010930348934729248,
      "loss": 1.1097,
      "step": 2530
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.13735607266426086,
      "learning_rate": 0.00010924143701712211,
      "loss": 1.1026,
      "step": 2531
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.17504572868347168,
      "learning_rate": 0.00010917938109768404,
      "loss": 1.1285,
      "step": 2532
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12810488045215607,
      "learning_rate": 0.00010911732161308015,
      "loss": 1.1237,
      "step": 2533
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1303734928369522,
      "learning_rate": 0.00010905525858741364,
      "loss": 1.0886,
      "step": 2534
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.15570834279060364,
      "learning_rate": 0.000108993192044789,
      "loss": 1.0912,
      "step": 2535
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12256300449371338,
      "learning_rate": 0.00010893112200931219,
      "loss": 1.0663,
      "step": 2536
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.13153968751430511,
      "learning_rate": 0.00010886904850509052,
      "loss": 1.1854,
      "step": 2537
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.15007434785366058,
      "learning_rate": 0.00010880697155623263,
      "loss": 1.1375,
      "step": 2538
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.15294648706912994,
      "learning_rate": 0.00010874489118684845,
      "loss": 1.1212,
      "step": 2539
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.11056141555309296,
      "learning_rate": 0.00010868280742104928,
      "loss": 1.1011,
      "step": 2540
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12624454498291016,
      "learning_rate": 0.00010862072028294776,
      "loss": 1.0813,
      "step": 2541
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1498328000307083,
      "learning_rate": 0.00010855862979665787,
      "loss": 1.0917,
      "step": 2542
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.132431760430336,
      "learning_rate": 0.00010849653598629476,
      "loss": 1.2619,
      "step": 2543
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.14327260851860046,
      "learning_rate": 0.00010843443887597495,
      "loss": 1.1421,
      "step": 2544
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.11624880880117416,
      "learning_rate": 0.00010837233848981632,
      "loss": 1.0916,
      "step": 2545
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.154770627617836,
      "learning_rate": 0.00010831023485193788,
      "loss": 1.1709,
      "step": 2546
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.09668345004320145,
      "learning_rate": 0.00010824812798645996,
      "loss": 1.189,
      "step": 2547
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.29019904136657715,
      "learning_rate": 0.00010818601791750417,
      "loss": 1.2415,
      "step": 2548
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.14910857379436493,
      "learning_rate": 0.00010812390466919338,
      "loss": 1.0717,
      "step": 2549
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.13996241986751556,
      "learning_rate": 0.00010806178826565161,
      "loss": 1.1487,
      "step": 2550
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.14915509521961212,
      "learning_rate": 0.00010799966873100417,
      "loss": 1.1714,
      "step": 2551
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1225634515285492,
      "learning_rate": 0.00010793754608937758,
      "loss": 1.1176,
      "step": 2552
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12608449161052704,
      "learning_rate": 0.00010787542036489955,
      "loss": 1.1039,
      "step": 2553
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.15166617929935455,
      "learning_rate": 0.00010781329158169902,
      "loss": 1.0942,
      "step": 2554
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.13943882286548615,
      "learning_rate": 0.00010775115976390606,
      "loss": 1.1182,
      "step": 2555
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1378665566444397,
      "learning_rate": 0.00010768902493565196,
      "loss": 1.2052,
      "step": 2556
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.13725712895393372,
      "learning_rate": 0.00010762688712106917,
      "loss": 1.1307,
      "step": 2557
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12152264267206192,
      "learning_rate": 0.00010756474634429132,
      "loss": 1.1441,
      "step": 2558
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.08334074914455414,
      "learning_rate": 0.00010750260262945314,
      "loss": 1.1627,
      "step": 2559
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.10767535865306854,
      "learning_rate": 0.00010744045600069055,
      "loss": 1.1231,
      "step": 2560
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.13590557873249054,
      "learning_rate": 0.00010737830648214062,
      "loss": 1.0908,
      "step": 2561
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.14964154362678528,
      "learning_rate": 0.00010731615409794143,
      "loss": 1.1626,
      "step": 2562
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1303054541349411,
      "learning_rate": 0.00010725399887223233,
      "loss": 1.0958,
      "step": 2563
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12171092629432678,
      "learning_rate": 0.00010719184082915363,
      "loss": 1.1179,
      "step": 2564
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.11213371157646179,
      "learning_rate": 0.00010712967999284682,
      "loss": 1.1166,
      "step": 2565
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.15331928431987762,
      "learning_rate": 0.00010706751638745447,
      "loss": 1.1325,
      "step": 2566
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.14010320603847504,
      "learning_rate": 0.00010700535003712022,
      "loss": 1.1397,
      "step": 2567
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.12474334239959717,
      "learning_rate": 0.0001069431809659887,
      "loss": 1.1586,
      "step": 2568
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.14930571615695953,
      "learning_rate": 0.0001068810091982057,
      "loss": 1.1151,
      "step": 2569
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.16670584678649902,
      "learning_rate": 0.00010681883475791803,
      "loss": 1.1171,
      "step": 2570
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.13524527847766876,
      "learning_rate": 0.0001067566576692735,
      "loss": 1.1007,
      "step": 2571
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.13782134652137756,
      "learning_rate": 0.00010669447795642103,
      "loss": 1.0411,
      "step": 2572
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1296006739139557,
      "learning_rate": 0.00010663229564351041,
      "loss": 1.1259,
      "step": 2573
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.15538984537124634,
      "learning_rate": 0.00010657011075469259,
      "loss": 1.1342,
      "step": 2574
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12411961704492569,
      "learning_rate": 0.0001065079233141195,
      "loss": 1.1543,
      "step": 2575
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1258893609046936,
      "learning_rate": 0.00010644573334594395,
      "loss": 1.1636,
      "step": 2576
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.10095099359750748,
      "learning_rate": 0.00010638354087431985,
      "loss": 1.2496,
      "step": 2577
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.16266852617263794,
      "learning_rate": 0.00010632134592340204,
      "loss": 1.12,
      "step": 2578
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1026599183678627,
      "learning_rate": 0.0001062591485173463,
      "loss": 1.0353,
      "step": 2579
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12140492349863052,
      "learning_rate": 0.00010619694868030942,
      "loss": 1.1282,
      "step": 2580
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11859510838985443,
      "learning_rate": 0.00010613474643644908,
      "loss": 1.1812,
      "step": 2581
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.10526355355978012,
      "learning_rate": 0.0001060725418099239,
      "loss": 1.1427,
      "step": 2582
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11631082743406296,
      "learning_rate": 0.00010601033482489346,
      "loss": 1.1432,
      "step": 2583
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13089729845523834,
      "learning_rate": 0.00010594812550551825,
      "loss": 1.1372,
      "step": 2584
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.10645223408937454,
      "learning_rate": 0.00010588591387595961,
      "loss": 1.1265,
      "step": 2585
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13449940085411072,
      "learning_rate": 0.00010582369996037984,
      "loss": 1.0862,
      "step": 2586
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12698443233966827,
      "learning_rate": 0.00010576148378294213,
      "loss": 1.1067,
      "step": 2587
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11834579706192017,
      "learning_rate": 0.0001056992653678105,
      "loss": 1.1043,
      "step": 2588
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.10321108996868134,
      "learning_rate": 0.00010563704473914986,
      "loss": 1.2041,
      "step": 2589
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1132575049996376,
      "learning_rate": 0.00010557482192112602,
      "loss": 1.1538,
      "step": 2590
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11528749763965607,
      "learning_rate": 0.00010551259693790556,
      "loss": 1.1984,
      "step": 2591
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1355505883693695,
      "learning_rate": 0.00010545036981365602,
      "loss": 1.1516,
      "step": 2592
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11153629422187805,
      "learning_rate": 0.0001053881405725456,
      "loss": 1.12,
      "step": 2593
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.14057233929634094,
      "learning_rate": 0.00010532590923874349,
      "loss": 1.1827,
      "step": 2594
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11393425613641739,
      "learning_rate": 0.00010526367583641958,
      "loss": 1.1839,
      "step": 2595
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1338927298784256,
      "learning_rate": 0.00010520144038974466,
      "loss": 1.0958,
      "step": 2596
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.157837375998497,
      "learning_rate": 0.00010513920292289021,
      "loss": 1.135,
      "step": 2597
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13308458030223846,
      "learning_rate": 0.00010507696346002858,
      "loss": 1.1572,
      "step": 2598
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.16041159629821777,
      "learning_rate": 0.00010501472202533284,
      "loss": 1.1404,
      "step": 2599
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1561173051595688,
      "learning_rate": 0.00010495247864297684,
      "loss": 1.0601,
      "step": 2600
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12204111367464066,
      "learning_rate": 0.00010489023333713521,
      "loss": 1.1363,
      "step": 2601
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11670318245887756,
      "learning_rate": 0.00010482798613198329,
      "loss": 1.1874,
      "step": 2602
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11118966341018677,
      "learning_rate": 0.00010476573705169719,
      "loss": 1.113,
      "step": 2603
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.15672335028648376,
      "learning_rate": 0.00010470348612045375,
      "loss": 1.1725,
      "step": 2604
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13462422788143158,
      "learning_rate": 0.00010464123336243048,
      "loss": 1.1469,
      "step": 2605
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12122746556997299,
      "learning_rate": 0.00010457897880180565,
      "loss": 1.1623,
      "step": 2606
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12082651257514954,
      "learning_rate": 0.00010451672246275825,
      "loss": 1.1349,
      "step": 2607
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13625676929950714,
      "learning_rate": 0.00010445446436946789,
      "loss": 1.0783,
      "step": 2608
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12511110305786133,
      "learning_rate": 0.00010439220454611487,
      "loss": 1.0881,
      "step": 2609
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12738360464572906,
      "learning_rate": 0.00010432994301688021,
      "loss": 1.0721,
      "step": 2610
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13192635774612427,
      "learning_rate": 0.00010426767980594558,
      "loss": 1.1149,
      "step": 2611
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13909798860549927,
      "learning_rate": 0.00010420541493749332,
      "loss": 1.1474,
      "step": 2612
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12073148041963577,
      "learning_rate": 0.00010414314843570634,
      "loss": 1.0724,
      "step": 2613
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1280267834663391,
      "learning_rate": 0.00010408088032476822,
      "loss": 1.0882,
      "step": 2614
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.12195128947496414,
      "learning_rate": 0.00010401861062886325,
      "loss": 1.0532,
      "step": 2615
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11220568418502808,
      "learning_rate": 0.00010395633937217621,
      "loss": 1.226,
      "step": 2616
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13371863961219788,
      "learning_rate": 0.00010389406657889254,
      "loss": 1.1492,
      "step": 2617
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1309518665075302,
      "learning_rate": 0.00010383179227319826,
      "loss": 1.1069,
      "step": 2618
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13498713076114655,
      "learning_rate": 0.00010376951647928005,
      "loss": 1.0522,
      "step": 2619
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.146668940782547,
      "learning_rate": 0.00010370723922132506,
      "loss": 1.124,
      "step": 2620
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.11095579713582993,
      "learning_rate": 0.00010364496052352108,
      "loss": 1.0951,
      "step": 2621
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1308859884738922,
      "learning_rate": 0.00010358268041005643,
      "loss": 1.103,
      "step": 2622
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13163278996944427,
      "learning_rate": 0.00010352039890511998,
      "loss": 1.1958,
      "step": 2623
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13302432000637054,
      "learning_rate": 0.00010345811603290118,
      "loss": 1.1915,
      "step": 2624
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13689318299293518,
      "learning_rate": 0.00010339583181758996,
      "loss": 1.126,
      "step": 2625
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.14390116930007935,
      "learning_rate": 0.00010333354628337679,
      "loss": 1.2134,
      "step": 2626
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12444696575403214,
      "learning_rate": 0.00010327125945445264,
      "loss": 1.098,
      "step": 2627
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1147523894906044,
      "learning_rate": 0.00010320897135500905,
      "loss": 1.1721,
      "step": 2628
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1303773671388626,
      "learning_rate": 0.00010314668200923792,
      "loss": 1.1708,
      "step": 2629
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.17957907915115356,
      "learning_rate": 0.00010308439144133177,
      "loss": 1.2224,
      "step": 2630
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.14394539594650269,
      "learning_rate": 0.00010302209967548353,
      "loss": 1.1193,
      "step": 2631
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12190341204404831,
      "learning_rate": 0.00010295980673588659,
      "loss": 1.1552,
      "step": 2632
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.15377764403820038,
      "learning_rate": 0.00010289751264673484,
      "loss": 1.1461,
      "step": 2633
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12752403318881989,
      "learning_rate": 0.00010283521743222255,
      "loss": 1.1173,
      "step": 2634
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.11877784878015518,
      "learning_rate": 0.00010277292111654446,
      "loss": 1.0839,
      "step": 2635
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.10796346515417099,
      "learning_rate": 0.00010271062372389581,
      "loss": 1.0885,
      "step": 2636
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12022894620895386,
      "learning_rate": 0.00010264832527847212,
      "loss": 1.1564,
      "step": 2637
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.11078213900327682,
      "learning_rate": 0.00010258602580446941,
      "loss": 1.1125,
      "step": 2638
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13750986754894257,
      "learning_rate": 0.00010252372532608405,
      "loss": 1.238,
      "step": 2639
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13050401210784912,
      "learning_rate": 0.0001024614238675129,
      "loss": 1.1556,
      "step": 2640
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12122219800949097,
      "learning_rate": 0.00010239912145295303,
      "loss": 1.1053,
      "step": 2641
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13207361102104187,
      "learning_rate": 0.00010233681810660207,
      "loss": 1.1529,
      "step": 2642
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.14329510927200317,
      "learning_rate": 0.00010227451385265787,
      "loss": 1.2168,
      "step": 2643
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.19093473255634308,
      "learning_rate": 0.00010221220871531869,
      "loss": 1.1081,
      "step": 2644
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.14228728413581848,
      "learning_rate": 0.00010214990271878318,
      "loss": 1.0892,
      "step": 2645
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.09503629058599472,
      "learning_rate": 0.00010208759588725016,
      "loss": 1.1032,
      "step": 2646
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1887892633676529,
      "learning_rate": 0.000102025288244919,
      "loss": 1.1199,
      "step": 2647
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13557374477386475,
      "learning_rate": 0.00010196297981598922,
      "loss": 1.1201,
      "step": 2648
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1065739169716835,
      "learning_rate": 0.00010190067062466069,
      "loss": 1.0754,
      "step": 2649
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.18160699307918549,
      "learning_rate": 0.00010183836069513359,
      "loss": 1.1678,
      "step": 2650
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13682709634304047,
      "learning_rate": 0.00010177605005160838,
      "loss": 1.1769,
      "step": 2651
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12422839552164078,
      "learning_rate": 0.00010171373871828578,
      "loss": 1.1429,
      "step": 2652
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.10330329090356827,
      "learning_rate": 0.00010165142671936685,
      "loss": 1.1245,
      "step": 2653
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1429406851530075,
      "learning_rate": 0.00010158911407905278,
      "loss": 1.1441,
      "step": 2654
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.11314301192760468,
      "learning_rate": 0.00010152680082154513,
      "loss": 1.0248,
      "step": 2655
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1296716332435608,
      "learning_rate": 0.0001014644869710456,
      "loss": 1.1721,
      "step": 2656
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1258556842803955,
      "learning_rate": 0.00010140217255175625,
      "loss": 1.0866,
      "step": 2657
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.15776795148849487,
      "learning_rate": 0.00010133985758787921,
      "loss": 1.1163,
      "step": 2658
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12838485836982727,
      "learning_rate": 0.00010127754210361693,
      "loss": 1.1494,
      "step": 2659
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.14994895458221436,
      "learning_rate": 0.00010121522612317203,
      "loss": 1.0963,
      "step": 2660
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.14245429635047913,
      "learning_rate": 0.00010115290967074728,
      "loss": 1.1411,
      "step": 2661
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.11902107298374176,
      "learning_rate": 0.00010109059277054574,
      "loss": 1.1049,
      "step": 2662
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12454280257225037,
      "learning_rate": 0.00010102827544677049,
      "loss": 1.0004,
      "step": 2663
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12659010291099548,
      "learning_rate": 0.00010096595772362491,
      "loss": 1.1816,
      "step": 2664
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.11207392066717148,
      "learning_rate": 0.0001009036396253125,
      "loss": 1.1413,
      "step": 2665
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13760559260845184,
      "learning_rate": 0.00010084132117603689,
      "loss": 1.1554,
      "step": 2666
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13545608520507812,
      "learning_rate": 0.00010077900240000179,
      "loss": 1.2825,
      "step": 2667
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13373397290706635,
      "learning_rate": 0.00010071668332141115,
      "loss": 1.107,
      "step": 2668
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.14539578557014465,
      "learning_rate": 0.00010065436396446899,
      "loss": 1.1467,
      "step": 2669
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.14728428423404694,
      "learning_rate": 0.00010059204435337937,
      "loss": 1.1915,
      "step": 2670
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.13219614326953888,
      "learning_rate": 0.00010052972451234657,
      "loss": 1.12,
      "step": 2671
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1018829420208931,
      "learning_rate": 0.00010046740446557486,
      "loss": 1.1034,
      "step": 2672
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.11733624339103699,
      "learning_rate": 0.00010040508423726865,
      "loss": 1.1652,
      "step": 2673
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.15831546485424042,
      "learning_rate": 0.00010034276385163237,
      "loss": 1.1063,
      "step": 2674
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.1309439241886139,
      "learning_rate": 0.00010028044333287057,
      "loss": 1.1,
      "step": 2675
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2509126663208008,
      "learning_rate": 0.0001002181227051878,
      "loss": 1.1816,
      "step": 2676
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12457872182130814,
      "learning_rate": 0.00010015580199278873,
      "loss": 1.1631,
      "step": 2677
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.14165012538433075,
      "learning_rate": 0.00010009348121987794,
      "loss": 1.1723,
      "step": 2678
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.14512354135513306,
      "learning_rate": 0.0001000311604106601,
      "loss": 1.1485,
      "step": 2679
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12050214409828186,
      "learning_rate": 9.996883958933992e-05,
      "loss": 1.1257,
      "step": 2680
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.13448520004749298,
      "learning_rate": 9.990651878012211e-05,
      "loss": 1.0586,
      "step": 2681
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11844149976968765,
      "learning_rate": 9.984419800721132e-05,
      "loss": 1.1951,
      "step": 2682
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12642155587673187,
      "learning_rate": 9.97818772948122e-05,
      "loss": 1.1079,
      "step": 2683
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.15310771763324738,
      "learning_rate": 9.971955666712944e-05,
      "loss": 1.1543,
      "step": 2684
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1282517910003662,
      "learning_rate": 9.965723614836764e-05,
      "loss": 1.0515,
      "step": 2685
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1483536809682846,
      "learning_rate": 9.959491576273139e-05,
      "loss": 1.1323,
      "step": 2686
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11740046739578247,
      "learning_rate": 9.953259553442518e-05,
      "loss": 1.1243,
      "step": 2687
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.10445234179496765,
      "learning_rate": 9.947027548765348e-05,
      "loss": 1.1949,
      "step": 2688
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1418544352054596,
      "learning_rate": 9.940795564662063e-05,
      "loss": 1.1117,
      "step": 2689
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1341327279806137,
      "learning_rate": 9.934563603553102e-05,
      "loss": 1.1014,
      "step": 2690
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.13397547602653503,
      "learning_rate": 9.928331667858886e-05,
      "loss": 1.1766,
      "step": 2691
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.14524011313915253,
      "learning_rate": 9.922099759999822e-05,
      "loss": 1.0952,
      "step": 2692
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1385146975517273,
      "learning_rate": 9.915867882396315e-05,
      "loss": 1.1131,
      "step": 2693
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.0908433347940445,
      "learning_rate": 9.909636037468753e-05,
      "loss": 1.3367,
      "step": 2694
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12240304797887802,
      "learning_rate": 9.903404227637508e-05,
      "loss": 1.0393,
      "step": 2695
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.14525730907917023,
      "learning_rate": 9.897172455322953e-05,
      "loss": 1.1075,
      "step": 2696
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11152852326631546,
      "learning_rate": 9.890940722945428e-05,
      "loss": 1.1126,
      "step": 2697
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12889206409454346,
      "learning_rate": 9.884709032925273e-05,
      "loss": 1.1719,
      "step": 2698
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.14006265997886658,
      "learning_rate": 9.878477387682803e-05,
      "loss": 1.1548,
      "step": 2699
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11841311305761337,
      "learning_rate": 9.872245789638308e-05,
      "loss": 1.0823,
      "step": 2700
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12788690626621246,
      "learning_rate": 9.866014241212079e-05,
      "loss": 1.1512,
      "step": 2701
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12002198398113251,
      "learning_rate": 9.859782744824376e-05,
      "loss": 1.1238,
      "step": 2702
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11864881217479706,
      "learning_rate": 9.85355130289544e-05,
      "loss": 1.1345,
      "step": 2703
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11675522476434708,
      "learning_rate": 9.84731991784549e-05,
      "loss": 1.1533,
      "step": 2704
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.18040168285369873,
      "learning_rate": 9.841088592094725e-05,
      "loss": 1.0647,
      "step": 2705
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11699610948562622,
      "learning_rate": 9.834857328063316e-05,
      "loss": 1.0985,
      "step": 2706
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1437186747789383,
      "learning_rate": 9.828626128171421e-05,
      "loss": 1.1105,
      "step": 2707
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1395156979560852,
      "learning_rate": 9.822394994839165e-05,
      "loss": 1.1502,
      "step": 2708
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11541968584060669,
      "learning_rate": 9.816163930486642e-05,
      "loss": 1.0924,
      "step": 2709
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1521672010421753,
      "learning_rate": 9.809932937533934e-05,
      "loss": 1.157,
      "step": 2710
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.09097780287265778,
      "learning_rate": 9.803702018401083e-05,
      "loss": 1.0825,
      "step": 2711
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12570732831954956,
      "learning_rate": 9.7974711755081e-05,
      "loss": 1.2349,
      "step": 2712
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11021190881729126,
      "learning_rate": 9.791240411274982e-05,
      "loss": 1.0148,
      "step": 2713
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12459529936313629,
      "learning_rate": 9.785009728121686e-05,
      "loss": 1.1006,
      "step": 2714
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1318003237247467,
      "learning_rate": 9.778779128468132e-05,
      "loss": 1.0177,
      "step": 2715
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.15788321197032928,
      "learning_rate": 9.772548614734217e-05,
      "loss": 1.1482,
      "step": 2716
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.13140280544757843,
      "learning_rate": 9.766318189339797e-05,
      "loss": 1.1355,
      "step": 2717
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.11584104597568512,
      "learning_rate": 9.760087854704698e-05,
      "loss": 1.126,
      "step": 2718
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.13363659381866455,
      "learning_rate": 9.753857613248714e-05,
      "loss": 1.1527,
      "step": 2719
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12988446652889252,
      "learning_rate": 9.747627467391596e-05,
      "loss": 1.1152,
      "step": 2720
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.17212513089179993,
      "learning_rate": 9.741397419553063e-05,
      "loss": 1.0604,
      "step": 2721
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1274261474609375,
      "learning_rate": 9.735167472152791e-05,
      "loss": 1.1604,
      "step": 2722
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1415371149778366,
      "learning_rate": 9.728937627610424e-05,
      "loss": 1.0987,
      "step": 2723
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1470460742712021,
      "learning_rate": 9.722707888345552e-05,
      "loss": 1.1012,
      "step": 2724
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.1276165097951889,
      "learning_rate": 9.716478256777747e-05,
      "loss": 1.1001,
      "step": 2725
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12065137177705765,
      "learning_rate": 9.710248735326519e-05,
      "loss": 1.1012,
      "step": 2726
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.12305580079555511,
      "learning_rate": 9.704019326411343e-05,
      "loss": 1.1627,
      "step": 2727
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.15387901663780212,
      "learning_rate": 9.697790032451651e-05,
      "loss": 1.2503,
      "step": 2728
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.08129476010799408,
      "learning_rate": 9.691560855866826e-05,
      "loss": 1.1401,
      "step": 2729
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.16048277914524078,
      "learning_rate": 9.685331799076209e-05,
      "loss": 1.1195,
      "step": 2730
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.12451910227537155,
      "learning_rate": 9.679102864499099e-05,
      "loss": 1.2183,
      "step": 2731
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1197393462061882,
      "learning_rate": 9.672874054554738e-05,
      "loss": 1.1064,
      "step": 2732
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.11675605177879333,
      "learning_rate": 9.666645371662324e-05,
      "loss": 1.1541,
      "step": 2733
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1492442935705185,
      "learning_rate": 9.660416818241007e-05,
      "loss": 1.108,
      "step": 2734
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.15328697860240936,
      "learning_rate": 9.654188396709882e-05,
      "loss": 1.0892,
      "step": 2735
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.16454927623271942,
      "learning_rate": 9.647960109488003e-05,
      "loss": 1.1003,
      "step": 2736
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1338902860879898,
      "learning_rate": 9.641731958994359e-05,
      "loss": 1.0964,
      "step": 2737
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.12653954327106476,
      "learning_rate": 9.635503947647894e-05,
      "loss": 1.1512,
      "step": 2738
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1473674178123474,
      "learning_rate": 9.629276077867497e-05,
      "loss": 1.2173,
      "step": 2739
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.12847444415092468,
      "learning_rate": 9.623048352071998e-05,
      "loss": 1.1179,
      "step": 2740
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.15160344541072845,
      "learning_rate": 9.616820772680173e-05,
      "loss": 1.1514,
      "step": 2741
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1416863054037094,
      "learning_rate": 9.610593342110746e-05,
      "loss": 1.0595,
      "step": 2742
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.14245428144931793,
      "learning_rate": 9.604366062782381e-05,
      "loss": 1.0974,
      "step": 2743
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.15476500988006592,
      "learning_rate": 9.598138937113676e-05,
      "loss": 1.1574,
      "step": 2744
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.128591388463974,
      "learning_rate": 9.591911967523179e-05,
      "loss": 1.1025,
      "step": 2745
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.13253438472747803,
      "learning_rate": 9.585685156429369e-05,
      "loss": 1.1797,
      "step": 2746
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.13782213628292084,
      "learning_rate": 9.579458506250669e-05,
      "loss": 1.0788,
      "step": 2747
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1393997073173523,
      "learning_rate": 9.573232019405441e-05,
      "loss": 1.0791,
      "step": 2748
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1360732614994049,
      "learning_rate": 9.567005698311981e-05,
      "loss": 1.1102,
      "step": 2749
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1318478286266327,
      "learning_rate": 9.560779545388517e-05,
      "loss": 1.1629,
      "step": 2750
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.14311620593070984,
      "learning_rate": 9.554553563053215e-05,
      "loss": 1.0963,
      "step": 2751
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.12694762647151947,
      "learning_rate": 9.54832775372418e-05,
      "loss": 1.1743,
      "step": 2752
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.14445050060749054,
      "learning_rate": 9.542102119819434e-05,
      "loss": 1.1367,
      "step": 2753
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.11144652962684631,
      "learning_rate": 9.535876663756954e-05,
      "loss": 1.2256,
      "step": 2754
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1413375735282898,
      "learning_rate": 9.529651387954627e-05,
      "loss": 1.1084,
      "step": 2755
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.12779457867145538,
      "learning_rate": 9.523426294830284e-05,
      "loss": 1.136,
      "step": 2756
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1526232808828354,
      "learning_rate": 9.517201386801675e-05,
      "loss": 1.1416,
      "step": 2757
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1089874655008316,
      "learning_rate": 9.510976666286484e-05,
      "loss": 1.0949,
      "step": 2758
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1309816688299179,
      "learning_rate": 9.504752135702317e-05,
      "loss": 1.1669,
      "step": 2759
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.18116767704486847,
      "learning_rate": 9.498527797466718e-05,
      "loss": 1.0431,
      "step": 2760
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.14649949967861176,
      "learning_rate": 9.492303653997146e-05,
      "loss": 1.1454,
      "step": 2761
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.13096138834953308,
      "learning_rate": 9.48607970771098e-05,
      "loss": 1.2142,
      "step": 2762
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.16232828795909882,
      "learning_rate": 9.479855961025536e-05,
      "loss": 1.0692,
      "step": 2763
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.15436621010303497,
      "learning_rate": 9.473632416358044e-05,
      "loss": 1.0765,
      "step": 2764
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.15464545786380768,
      "learning_rate": 9.467409076125652e-05,
      "loss": 1.1144,
      "step": 2765
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.14814594388008118,
      "learning_rate": 9.461185942745443e-05,
      "loss": 1.2134,
      "step": 2766
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.14494243264198303,
      "learning_rate": 9.454963018634402e-05,
      "loss": 1.075,
      "step": 2767
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1293327808380127,
      "learning_rate": 9.448740306209446e-05,
      "loss": 1.1078,
      "step": 2768
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1165275126695633,
      "learning_rate": 9.442517807887403e-05,
      "loss": 1.0386,
      "step": 2769
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1289057731628418,
      "learning_rate": 9.436295526085015e-05,
      "loss": 1.0829,
      "step": 2770
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.16054515540599823,
      "learning_rate": 9.430073463218951e-05,
      "loss": 1.1472,
      "step": 2771
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.10987726598978043,
      "learning_rate": 9.42385162170579e-05,
      "loss": 1.1021,
      "step": 2772
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.16421402990818024,
      "learning_rate": 9.417630003962018e-05,
      "loss": 1.0817,
      "step": 2773
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.131307452917099,
      "learning_rate": 9.411408612404042e-05,
      "loss": 1.1213,
      "step": 2774
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.12413746118545532,
      "learning_rate": 9.405187449448179e-05,
      "loss": 1.1541,
      "step": 2775
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.11998411267995834,
      "learning_rate": 9.398966517510653e-05,
      "loss": 1.1308,
      "step": 2776
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.13310159742832184,
      "learning_rate": 9.392745819007611e-05,
      "loss": 1.1356,
      "step": 2777
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.15277954936027527,
      "learning_rate": 9.386525356355094e-05,
      "loss": 1.0794,
      "step": 2778
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1344311535358429,
      "learning_rate": 9.38030513196906e-05,
      "loss": 1.1607,
      "step": 2779
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.09580395370721817,
      "learning_rate": 9.37408514826537e-05,
      "loss": 1.03,
      "step": 2780
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.10243196785449982,
      "learning_rate": 9.3678654076598e-05,
      "loss": 1.1408,
      "step": 2781
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.18442301452159882,
      "learning_rate": 9.361645912568016e-05,
      "loss": 1.1342,
      "step": 2782
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.11750411987304688,
      "learning_rate": 9.355426665405606e-05,
      "loss": 1.1963,
      "step": 2783
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.12270501255989075,
      "learning_rate": 9.349207668588052e-05,
      "loss": 1.0476,
      "step": 2784
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1405157893896103,
      "learning_rate": 9.342988924530742e-05,
      "loss": 1.1538,
      "step": 2785
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1681136041879654,
      "learning_rate": 9.336770435648964e-05,
      "loss": 1.0524,
      "step": 2786
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.14537014067173004,
      "learning_rate": 9.330552204357904e-05,
      "loss": 1.1769,
      "step": 2787
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.12974117696285248,
      "learning_rate": 9.324334233072648e-05,
      "loss": 1.1532,
      "step": 2788
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.15788188576698303,
      "learning_rate": 9.318116524208198e-05,
      "loss": 1.1917,
      "step": 2789
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.12609392404556274,
      "learning_rate": 9.311899080179433e-05,
      "loss": 1.2075,
      "step": 2790
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.13059960305690765,
      "learning_rate": 9.305681903401132e-05,
      "loss": 1.1348,
      "step": 2791
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.13584473729133606,
      "learning_rate": 9.299464996287983e-05,
      "loss": 1.1974,
      "step": 2792
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.09251294285058975,
      "learning_rate": 9.293248361254556e-05,
      "loss": 1.0758,
      "step": 2793
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.13312803208827972,
      "learning_rate": 9.287032000715318e-05,
      "loss": 1.1836,
      "step": 2794
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.10717364400625229,
      "learning_rate": 9.28081591708464e-05,
      "loss": 1.1132,
      "step": 2795
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1289997696876526,
      "learning_rate": 9.274600112776769e-05,
      "loss": 1.168,
      "step": 2796
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.18605946004390717,
      "learning_rate": 9.268384590205858e-05,
      "loss": 1.133,
      "step": 2797
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.14424335956573486,
      "learning_rate": 9.262169351785943e-05,
      "loss": 1.1539,
      "step": 2798
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.14781324565410614,
      "learning_rate": 9.255954399930947e-05,
      "loss": 1.2053,
      "step": 2799
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.11874663084745407,
      "learning_rate": 9.249739737054686e-05,
      "loss": 1.0907,
      "step": 2800
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.13128086924552917,
      "learning_rate": 9.24352536557087e-05,
      "loss": 1.1031,
      "step": 2801
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1144358292222023,
      "learning_rate": 9.237311287893086e-05,
      "loss": 1.1443,
      "step": 2802
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1322230100631714,
      "learning_rate": 9.231097506434807e-05,
      "loss": 1.2147,
      "step": 2803
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1606566458940506,
      "learning_rate": 9.224884023609397e-05,
      "loss": 1.1499,
      "step": 2804
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.12518586218357086,
      "learning_rate": 9.218670841830098e-05,
      "loss": 1.12,
      "step": 2805
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.14768366515636444,
      "learning_rate": 9.212457963510044e-05,
      "loss": 1.0995,
      "step": 2806
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.14345091581344604,
      "learning_rate": 9.206245391062244e-05,
      "loss": 1.1268,
      "step": 2807
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1257777214050293,
      "learning_rate": 9.200033126899584e-05,
      "loss": 1.1632,
      "step": 2808
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.15238025784492493,
      "learning_rate": 9.193821173434842e-05,
      "loss": 1.1022,
      "step": 2809
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.11585281044244766,
      "learning_rate": 9.187609533080667e-05,
      "loss": 1.1844,
      "step": 2810
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.12110118567943573,
      "learning_rate": 9.181398208249583e-05,
      "loss": 1.2237,
      "step": 2811
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.17074714601039886,
      "learning_rate": 9.175187201354004e-05,
      "loss": 1.125,
      "step": 2812
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.09794455766677856,
      "learning_rate": 9.168976514806216e-05,
      "loss": 1.0103,
      "step": 2813
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.13278347253799438,
      "learning_rate": 9.162766151018372e-05,
      "loss": 1.1297,
      "step": 2814
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.12654097378253937,
      "learning_rate": 9.156556112402507e-05,
      "loss": 1.1096,
      "step": 2815
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1458834409713745,
      "learning_rate": 9.150346401370527e-05,
      "loss": 1.1115,
      "step": 2816
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.10799829661846161,
      "learning_rate": 9.144137020334214e-05,
      "loss": 1.0996,
      "step": 2817
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.13706466555595398,
      "learning_rate": 9.137927971705222e-05,
      "loss": 1.1868,
      "step": 2818
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1511416733264923,
      "learning_rate": 9.131719257895074e-05,
      "loss": 1.0899,
      "step": 2819
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.12125783413648605,
      "learning_rate": 9.125510881315158e-05,
      "loss": 1.1466,
      "step": 2820
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1475456953048706,
      "learning_rate": 9.119302844376741e-05,
      "loss": 1.105,
      "step": 2821
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1447562426328659,
      "learning_rate": 9.113095149490952e-05,
      "loss": 1.0771,
      "step": 2822
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1223553717136383,
      "learning_rate": 9.106887799068782e-05,
      "loss": 1.1513,
      "step": 2823
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1462927907705307,
      "learning_rate": 9.100680795521104e-05,
      "loss": 1.1142,
      "step": 2824
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.11355869472026825,
      "learning_rate": 9.09447414125864e-05,
      "loss": 1.1349,
      "step": 2825
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.10961100459098816,
      "learning_rate": 9.088267838691986e-05,
      "loss": 1.1358,
      "step": 2826
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.16772285103797913,
      "learning_rate": 9.0820618902316e-05,
      "loss": 1.0904,
      "step": 2827
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.12750592827796936,
      "learning_rate": 9.075856298287795e-05,
      "loss": 1.1192,
      "step": 2828
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.11275243759155273,
      "learning_rate": 9.069651065270752e-05,
      "loss": 1.1334,
      "step": 2829
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.11856981366872787,
      "learning_rate": 9.06344619359052e-05,
      "loss": 1.0917,
      "step": 2830
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.12617258727550507,
      "learning_rate": 9.057241685656995e-05,
      "loss": 1.2196,
      "step": 2831
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.1383393108844757,
      "learning_rate": 9.051037543879932e-05,
      "loss": 1.1154,
      "step": 2832
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1237589418888092,
      "learning_rate": 9.044833770668956e-05,
      "loss": 1.0989,
      "step": 2833
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.11824850738048553,
      "learning_rate": 9.038630368433535e-05,
      "loss": 1.1061,
      "step": 2834
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.15545810759067535,
      "learning_rate": 9.032427339583e-05,
      "loss": 1.2168,
      "step": 2835
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.09221438318490982,
      "learning_rate": 9.026224686526538e-05,
      "loss": 1.1612,
      "step": 2836
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.12509042024612427,
      "learning_rate": 9.020022411673187e-05,
      "loss": 1.0531,
      "step": 2837
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1131993904709816,
      "learning_rate": 9.01382051743184e-05,
      "loss": 1.0856,
      "step": 2838
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.15923179686069489,
      "learning_rate": 9.007619006211241e-05,
      "loss": 1.0479,
      "step": 2839
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1386692225933075,
      "learning_rate": 9.00141788041998e-05,
      "loss": 1.0796,
      "step": 2840
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.12901721894741058,
      "learning_rate": 8.99521714246651e-05,
      "loss": 1.2331,
      "step": 2841
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.13934755325317383,
      "learning_rate": 8.989016794759127e-05,
      "loss": 1.0975,
      "step": 2842
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1532430797815323,
      "learning_rate": 8.982816839705969e-05,
      "loss": 1.1852,
      "step": 2843
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.16195225715637207,
      "learning_rate": 8.976617279715031e-05,
      "loss": 1.149,
      "step": 2844
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1802736073732376,
      "learning_rate": 8.970418117194146e-05,
      "loss": 1.1037,
      "step": 2845
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.11968957632780075,
      "learning_rate": 8.964219354550999e-05,
      "loss": 1.1223,
      "step": 2846
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.14212122559547424,
      "learning_rate": 8.958020994193124e-05,
      "loss": 1.0885,
      "step": 2847
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.14212122559547424,
      "learning_rate": 8.958020994193124e-05,
      "loss": 1.1648,
      "step": 2848
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.14296545088291168,
      "learning_rate": 8.951823038527888e-05,
      "loss": 1.1049,
      "step": 2849
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1392725259065628,
      "learning_rate": 8.945625489962503e-05,
      "loss": 1.2091,
      "step": 2850
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.12358943372964859,
      "learning_rate": 8.93942835090403e-05,
      "loss": 1.121,
      "step": 2851
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.12814457714557648,
      "learning_rate": 8.933231623759365e-05,
      "loss": 1.1455,
      "step": 2852
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.12035926431417465,
      "learning_rate": 8.927035310935242e-05,
      "loss": 1.1502,
      "step": 2853
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.13544492423534393,
      "learning_rate": 8.920839414838242e-05,
      "loss": 1.1276,
      "step": 2854
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.141524076461792,
      "learning_rate": 8.914643937874778e-05,
      "loss": 1.1129,
      "step": 2855
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.12275050580501556,
      "learning_rate": 8.908448882451104e-05,
      "loss": 1.1279,
      "step": 2856
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.13490058481693268,
      "learning_rate": 8.902254250973305e-05,
      "loss": 1.182,
      "step": 2857
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.14908047020435333,
      "learning_rate": 8.896060045847304e-05,
      "loss": 1.1019,
      "step": 2858
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.13848187029361725,
      "learning_rate": 8.889866269478859e-05,
      "loss": 1.1739,
      "step": 2859
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.11221689730882645,
      "learning_rate": 8.883672924273566e-05,
      "loss": 1.1728,
      "step": 2860
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.11518759280443192,
      "learning_rate": 8.877480012636847e-05,
      "loss": 1.0941,
      "step": 2861
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1476537585258484,
      "learning_rate": 8.871287536973953e-05,
      "loss": 1.1126,
      "step": 2862
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1245974749326706,
      "learning_rate": 8.865095499689977e-05,
      "loss": 1.2077,
      "step": 2863
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.15348005294799805,
      "learning_rate": 8.858903903189831e-05,
      "loss": 1.0448,
      "step": 2864
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.13377365469932556,
      "learning_rate": 8.852712749878254e-05,
      "loss": 1.1825,
      "step": 2865
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.14473237097263336,
      "learning_rate": 8.846522042159832e-05,
      "loss": 1.1598,
      "step": 2866
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1115972250699997,
      "learning_rate": 8.840331782438953e-05,
      "loss": 1.2159,
      "step": 2867
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.18417808413505554,
      "learning_rate": 8.834141973119849e-05,
      "loss": 1.1753,
      "step": 2868
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1108471229672432,
      "learning_rate": 8.82795261660657e-05,
      "loss": 1.0505,
      "step": 2869
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1318518966436386,
      "learning_rate": 8.821763715302987e-05,
      "loss": 1.0788,
      "step": 2870
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1272166520357132,
      "learning_rate": 8.815575271612797e-05,
      "loss": 1.1246,
      "step": 2871
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.13672739267349243,
      "learning_rate": 8.809387287939528e-05,
      "loss": 1.0922,
      "step": 2872
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1222548559308052,
      "learning_rate": 8.803199766686516e-05,
      "loss": 1.0937,
      "step": 2873
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.13209053874015808,
      "learning_rate": 8.797012710256923e-05,
      "loss": 1.1341,
      "step": 2874
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.10917244106531143,
      "learning_rate": 8.790826121053733e-05,
      "loss": 1.1276,
      "step": 2875
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.13175739347934723,
      "learning_rate": 8.784640001479741e-05,
      "loss": 1.2062,
      "step": 2876
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1151314303278923,
      "learning_rate": 8.77845435393757e-05,
      "loss": 1.2059,
      "step": 2877
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.12439488619565964,
      "learning_rate": 8.772269180829653e-05,
      "loss": 1.0909,
      "step": 2878
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.12815432250499725,
      "learning_rate": 8.766084484558237e-05,
      "loss": 1.0949,
      "step": 2879
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.130171537399292,
      "learning_rate": 8.759900267525391e-05,
      "loss": 1.16,
      "step": 2880
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.13244511187076569,
      "learning_rate": 8.753716532132992e-05,
      "loss": 1.1458,
      "step": 2881
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1433570683002472,
      "learning_rate": 8.747533280782726e-05,
      "loss": 1.17,
      "step": 2882
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.1643097847700119,
      "learning_rate": 8.741350515876104e-05,
      "loss": 1.1015,
      "step": 2883
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.12619829177856445,
      "learning_rate": 8.735168239814439e-05,
      "loss": 1.1663,
      "step": 2884
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.09733550250530243,
      "learning_rate": 8.728986454998858e-05,
      "loss": 1.1806,
      "step": 2885
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.10959547758102417,
      "learning_rate": 8.72280516383029e-05,
      "loss": 1.0589,
      "step": 2886
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.10724354535341263,
      "learning_rate": 8.716624368709477e-05,
      "loss": 1.129,
      "step": 2887
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.13084174692630768,
      "learning_rate": 8.71044407203697e-05,
      "loss": 1.114,
      "step": 2888
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14227478206157684,
      "learning_rate": 8.704264276213129e-05,
      "loss": 1.1262,
      "step": 2889
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.10486651957035065,
      "learning_rate": 8.698084983638111e-05,
      "loss": 1.1288,
      "step": 2890
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.12913037836551666,
      "learning_rate": 8.691906196711884e-05,
      "loss": 1.1192,
      "step": 2891
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11931903660297394,
      "learning_rate": 8.685727917834217e-05,
      "loss": 1.1222,
      "step": 2892
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14422185719013214,
      "learning_rate": 8.679550149404685e-05,
      "loss": 1.0904,
      "step": 2893
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1275436133146286,
      "learning_rate": 8.673372893822654e-05,
      "loss": 1.0691,
      "step": 2894
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.143745556473732,
      "learning_rate": 8.667196153487307e-05,
      "loss": 1.1296,
      "step": 2895
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1446634829044342,
      "learning_rate": 8.661019930797614e-05,
      "loss": 1.0875,
      "step": 2896
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.10595100373029709,
      "learning_rate": 8.654844228152355e-05,
      "loss": 1.0873,
      "step": 2897
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14193002879619598,
      "learning_rate": 8.648669047950097e-05,
      "loss": 1.1442,
      "step": 2898
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.13120344281196594,
      "learning_rate": 8.642494392589206e-05,
      "loss": 1.1643,
      "step": 2899
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.12806907296180725,
      "learning_rate": 8.63632026446785e-05,
      "loss": 1.1233,
      "step": 2900
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11999775469303131,
      "learning_rate": 8.630146665983992e-05,
      "loss": 1.0424,
      "step": 2901
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11778610199689865,
      "learning_rate": 8.623973599535385e-05,
      "loss": 1.0849,
      "step": 2902
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.134682297706604,
      "learning_rate": 8.617801067519574e-05,
      "loss": 1.1999,
      "step": 2903
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11348096281290054,
      "learning_rate": 8.611629072333904e-05,
      "loss": 1.1139,
      "step": 2904
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1139945238828659,
      "learning_rate": 8.605457616375503e-05,
      "loss": 1.1636,
      "step": 2905
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.13193963468074799,
      "learning_rate": 8.599286702041292e-05,
      "loss": 1.1407,
      "step": 2906
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11833209544420242,
      "learning_rate": 8.593116331727987e-05,
      "loss": 1.2203,
      "step": 2907
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.19956596195697784,
      "learning_rate": 8.586946507832088e-05,
      "loss": 1.1184,
      "step": 2908
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1642429381608963,
      "learning_rate": 8.580777232749883e-05,
      "loss": 1.1278,
      "step": 2909
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.12820953130722046,
      "learning_rate": 8.574608508877448e-05,
      "loss": 1.1591,
      "step": 2910
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.10999955981969833,
      "learning_rate": 8.568440338610638e-05,
      "loss": 1.1069,
      "step": 2911
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14158648252487183,
      "learning_rate": 8.562272724345108e-05,
      "loss": 1.0843,
      "step": 2912
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1266534924507141,
      "learning_rate": 8.556105668476286e-05,
      "loss": 1.166,
      "step": 2913
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.13802991807460785,
      "learning_rate": 8.549939173399385e-05,
      "loss": 1.0951,
      "step": 2914
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.18141940236091614,
      "learning_rate": 8.5437732415094e-05,
      "loss": 1.1919,
      "step": 2915
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11884453892707825,
      "learning_rate": 8.537607875201106e-05,
      "loss": 1.1156,
      "step": 2916
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11325988173484802,
      "learning_rate": 8.531443076869059e-05,
      "loss": 1.1461,
      "step": 2917
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.16384954750537872,
      "learning_rate": 8.525278848907604e-05,
      "loss": 1.1662,
      "step": 2918
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14672419428825378,
      "learning_rate": 8.519115193710849e-05,
      "loss": 1.164,
      "step": 2919
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.13492940366268158,
      "learning_rate": 8.512952113672689e-05,
      "loss": 1.1421,
      "step": 2920
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14208056032657623,
      "learning_rate": 8.506789611186794e-05,
      "loss": 1.1018,
      "step": 2921
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.10829965025186539,
      "learning_rate": 8.500627688646607e-05,
      "loss": 1.1112,
      "step": 2922
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.0920204222202301,
      "learning_rate": 8.494466348445345e-05,
      "loss": 1.1216,
      "step": 2923
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11137975752353668,
      "learning_rate": 8.48830559297601e-05,
      "loss": 1.1196,
      "step": 2924
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.13024303317070007,
      "learning_rate": 8.48214542463136e-05,
      "loss": 1.1506,
      "step": 2925
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14299383759498596,
      "learning_rate": 8.475985845803939e-05,
      "loss": 1.136,
      "step": 2926
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1487317830324173,
      "learning_rate": 8.469826858886054e-05,
      "loss": 1.1068,
      "step": 2927
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1684814840555191,
      "learning_rate": 8.463668466269784e-05,
      "loss": 1.1235,
      "step": 2928
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.13260984420776367,
      "learning_rate": 8.457510670346976e-05,
      "loss": 1.1228,
      "step": 2929
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.13194327056407928,
      "learning_rate": 8.451353473509253e-05,
      "loss": 1.1177,
      "step": 2930
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1414726972579956,
      "learning_rate": 8.445196878147996e-05,
      "loss": 1.1554,
      "step": 2931
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.12979888916015625,
      "learning_rate": 8.439040886654355e-05,
      "loss": 1.1017,
      "step": 2932
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.10439146310091019,
      "learning_rate": 8.432885501419247e-05,
      "loss": 1.0588,
      "step": 2933
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.11701620370149612,
      "learning_rate": 8.426730724833354e-05,
      "loss": 1.1286,
      "step": 2934
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1329582929611206,
      "learning_rate": 8.420576559287112e-05,
      "loss": 1.108,
      "step": 2935
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.105800561606884,
      "learning_rate": 8.414423007170741e-05,
      "loss": 1.0671,
      "step": 2936
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.12399829179048538,
      "learning_rate": 8.4082700708742e-05,
      "loss": 1.2336,
      "step": 2937
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.14813688397407532,
      "learning_rate": 8.402117752787226e-05,
      "loss": 1.1039,
      "step": 2938
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.09186362475156784,
      "learning_rate": 8.395966055299303e-05,
      "loss": 1.123,
      "step": 2939
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.12279273569583893,
      "learning_rate": 8.389814980799678e-05,
      "loss": 1.1247,
      "step": 2940
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.13048605620861053,
      "learning_rate": 8.38366453167736e-05,
      "loss": 1.1785,
      "step": 2941
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.12568090856075287,
      "learning_rate": 8.377514710321117e-05,
      "loss": 1.1177,
      "step": 2942
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.15872572362422943,
      "learning_rate": 8.371365519119462e-05,
      "loss": 1.1195,
      "step": 2943
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.15559762716293335,
      "learning_rate": 8.365216960460676e-05,
      "loss": 1.1159,
      "step": 2944
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.14286285638809204,
      "learning_rate": 8.35906903673278e-05,
      "loss": 1.1558,
      "step": 2945
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.14312584698200226,
      "learning_rate": 8.352921750323562e-05,
      "loss": 1.0454,
      "step": 2946
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1595708578824997,
      "learning_rate": 8.346775103620559e-05,
      "loss": 1.0931,
      "step": 2947
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.11477573961019516,
      "learning_rate": 8.340629099011057e-05,
      "loss": 1.213,
      "step": 2948
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.11624598503112793,
      "learning_rate": 8.334483738882089e-05,
      "loss": 1.0237,
      "step": 2949
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.12525023519992828,
      "learning_rate": 8.328339025620449e-05,
      "loss": 1.1077,
      "step": 2950
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.16388247907161713,
      "learning_rate": 8.322194961612667e-05,
      "loss": 1.3402,
      "step": 2951
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1418866664171219,
      "learning_rate": 8.316051549245025e-05,
      "loss": 1.0626,
      "step": 2952
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.13605566322803497,
      "learning_rate": 8.309908790903561e-05,
      "loss": 1.0859,
      "step": 2953
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.15496669709682465,
      "learning_rate": 8.303766688974046e-05,
      "loss": 1.0705,
      "step": 2954
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.14511381089687347,
      "learning_rate": 8.297625245842005e-05,
      "loss": 1.1302,
      "step": 2955
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1314726620912552,
      "learning_rate": 8.291484463892702e-05,
      "loss": 1.1351,
      "step": 2956
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.11887793242931366,
      "learning_rate": 8.285344345511146e-05,
      "loss": 1.2015,
      "step": 2957
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.11681418865919113,
      "learning_rate": 8.279204893082084e-05,
      "loss": 1.1607,
      "step": 2958
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.12796790897846222,
      "learning_rate": 8.273066108990018e-05,
      "loss": 1.146,
      "step": 2959
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.15571115911006927,
      "learning_rate": 8.266927995619174e-05,
      "loss": 1.0436,
      "step": 2960
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.21564966440200806,
      "learning_rate": 8.260790555353526e-05,
      "loss": 1.1075,
      "step": 2961
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.19380514323711395,
      "learning_rate": 8.254653790576787e-05,
      "loss": 1.0567,
      "step": 2962
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.12809507548809052,
      "learning_rate": 8.248517703672403e-05,
      "loss": 1.1336,
      "step": 2963
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1485135704278946,
      "learning_rate": 8.242382297023559e-05,
      "loss": 1.2055,
      "step": 2964
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1253761202096939,
      "learning_rate": 8.23624757301318e-05,
      "loss": 1.1843,
      "step": 2965
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.12173789739608765,
      "learning_rate": 8.230113534023918e-05,
      "loss": 1.0652,
      "step": 2966
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.17208485305309296,
      "learning_rate": 8.223980182438167e-05,
      "loss": 1.0898,
      "step": 2967
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.15911251306533813,
      "learning_rate": 8.217847520638048e-05,
      "loss": 1.0955,
      "step": 2968
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.18216827511787415,
      "learning_rate": 8.211715551005415e-05,
      "loss": 1.1684,
      "step": 2969
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1359216272830963,
      "learning_rate": 8.205584275921854e-05,
      "loss": 1.1913,
      "step": 2970
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.15732519328594208,
      "learning_rate": 8.199453697768686e-05,
      "loss": 1.187,
      "step": 2971
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1409125030040741,
      "learning_rate": 8.193323818926954e-05,
      "loss": 1.1689,
      "step": 2972
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.13141357898712158,
      "learning_rate": 8.187194641777431e-05,
      "loss": 1.1171,
      "step": 2973
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.13761527836322784,
      "learning_rate": 8.181066168700622e-05,
      "loss": 1.0556,
      "step": 2974
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.11938966065645218,
      "learning_rate": 8.174938402076754e-05,
      "loss": 1.1037,
      "step": 2975
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.14128342270851135,
      "learning_rate": 8.168811344285776e-05,
      "loss": 1.1276,
      "step": 2976
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.11739902198314667,
      "learning_rate": 8.162684997707374e-05,
      "loss": 1.1886,
      "step": 2977
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.14332614839076996,
      "learning_rate": 8.156559364720947e-05,
      "loss": 1.1045,
      "step": 2978
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.13234393298625946,
      "learning_rate": 8.150434447705623e-05,
      "loss": 1.1598,
      "step": 2979
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.13143517076969147,
      "learning_rate": 8.144310249040246e-05,
      "loss": 1.1674,
      "step": 2980
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.12094447016716003,
      "learning_rate": 8.138186771103382e-05,
      "loss": 1.0678,
      "step": 2981
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.11501040309667587,
      "learning_rate": 8.132064016273324e-05,
      "loss": 1.0761,
      "step": 2982
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.10691485553979874,
      "learning_rate": 8.12594198692808e-05,
      "loss": 1.1803,
      "step": 2983
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.17297294735908508,
      "learning_rate": 8.119820685445372e-05,
      "loss": 1.2223,
      "step": 2984
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1414119005203247,
      "learning_rate": 8.113700114202648e-05,
      "loss": 1.0045,
      "step": 2985
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.13404275476932526,
      "learning_rate": 8.107580275577058e-05,
      "loss": 1.1818,
      "step": 2986
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1350497454404831,
      "learning_rate": 8.101461171945483e-05,
      "loss": 1.2112,
      "step": 2987
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.11686667054891586,
      "learning_rate": 8.095342805684515e-05,
      "loss": 1.1521,
      "step": 2988
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12027668207883835,
      "learning_rate": 8.089225179170455e-05,
      "loss": 1.1354,
      "step": 2989
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1216755136847496,
      "learning_rate": 8.083108294779314e-05,
      "loss": 1.1042,
      "step": 2990
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12720739841461182,
      "learning_rate": 8.076992154886826e-05,
      "loss": 1.1189,
      "step": 2991
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.13684779405593872,
      "learning_rate": 8.070876761868426e-05,
      "loss": 1.211,
      "step": 2992
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1508704572916031,
      "learning_rate": 8.064762118099258e-05,
      "loss": 1.082,
      "step": 2993
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.10878854244947433,
      "learning_rate": 8.058648225954188e-05,
      "loss": 1.2002,
      "step": 2994
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.16376781463623047,
      "learning_rate": 8.052535087807774e-05,
      "loss": 1.1877,
      "step": 2995
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.18429052829742432,
      "learning_rate": 8.046422706034293e-05,
      "loss": 1.1415,
      "step": 2996
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1515297293663025,
      "learning_rate": 8.040311083007725e-05,
      "loss": 1.1117,
      "step": 2997
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.3754041790962219,
      "learning_rate": 8.034200221101746e-05,
      "loss": 1.0962,
      "step": 2998
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12820734083652496,
      "learning_rate": 8.028090122689748e-05,
      "loss": 1.1915,
      "step": 2999
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.11281125992536545,
      "learning_rate": 8.021980790144827e-05,
      "loss": 1.1656,
      "step": 3000
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1316874623298645,
      "learning_rate": 8.015872225839776e-05,
      "loss": 1.1903,
      "step": 3001
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.13510164618492126,
      "learning_rate": 8.009764432147087e-05,
      "loss": 1.2117,
      "step": 3002
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.11399060487747192,
      "learning_rate": 8.00365741143896e-05,
      "loss": 1.096,
      "step": 3003
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.11505240947008133,
      "learning_rate": 7.997551166087292e-05,
      "loss": 1.1176,
      "step": 3004
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.16501560807228088,
      "learning_rate": 7.991445698463672e-05,
      "loss": 1.0979,
      "step": 3005
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.10522349178791046,
      "learning_rate": 7.985341010939402e-05,
      "loss": 1.1822,
      "step": 3006
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.11124224215745926,
      "learning_rate": 7.979237105885467e-05,
      "loss": 1.1431,
      "step": 3007
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12357372045516968,
      "learning_rate": 7.973133985672558e-05,
      "loss": 1.0754,
      "step": 3008
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12721125781536102,
      "learning_rate": 7.967031652671051e-05,
      "loss": 1.1635,
      "step": 3009
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1321125477552414,
      "learning_rate": 7.960930109251023e-05,
      "loss": 1.0943,
      "step": 3010
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.11495645344257355,
      "learning_rate": 7.954829357782243e-05,
      "loss": 1.1287,
      "step": 3011
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1539977490901947,
      "learning_rate": 7.948729400634178e-05,
      "loss": 1.0763,
      "step": 3012
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.10956403613090515,
      "learning_rate": 7.942630240175977e-05,
      "loss": 1.1122,
      "step": 3013
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12778769433498383,
      "learning_rate": 7.936531878776483e-05,
      "loss": 1.1033,
      "step": 3014
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.13361407816410065,
      "learning_rate": 7.930434318804229e-05,
      "loss": 1.1953,
      "step": 3015
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12211057543754578,
      "learning_rate": 7.924337562627435e-05,
      "loss": 1.1737,
      "step": 3016
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.10309911519289017,
      "learning_rate": 7.918241612614017e-05,
      "loss": 1.1278,
      "step": 3017
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.17102864384651184,
      "learning_rate": 7.91214647113157e-05,
      "loss": 1.1444,
      "step": 3018
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.29254963994026184,
      "learning_rate": 7.906052140547373e-05,
      "loss": 1.1348,
      "step": 3019
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.16751250624656677,
      "learning_rate": 7.899958623228397e-05,
      "loss": 1.1449,
      "step": 3020
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.11628375947475433,
      "learning_rate": 7.893865921541295e-05,
      "loss": 1.1868,
      "step": 3021
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.14764338731765747,
      "learning_rate": 7.887774037852395e-05,
      "loss": 1.1679,
      "step": 3022
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.11825661361217499,
      "learning_rate": 7.881682974527723e-05,
      "loss": 1.2397,
      "step": 3023
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.16001632809638977,
      "learning_rate": 7.875592733932973e-05,
      "loss": 1.2611,
      "step": 3024
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.17567534744739532,
      "learning_rate": 7.869503318433528e-05,
      "loss": 1.1734,
      "step": 3025
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12159188836812973,
      "learning_rate": 7.863414730394443e-05,
      "loss": 1.048,
      "step": 3026
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12456348538398743,
      "learning_rate": 7.857326972180455e-05,
      "loss": 1.2051,
      "step": 3027
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.13221926987171173,
      "learning_rate": 7.85124004615598e-05,
      "loss": 1.1773,
      "step": 3028
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1514803022146225,
      "learning_rate": 7.845153954685115e-05,
      "loss": 1.0986,
      "step": 3029
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.18992994725704193,
      "learning_rate": 7.839068700131623e-05,
      "loss": 1.0959,
      "step": 3030
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12055522948503494,
      "learning_rate": 7.832984284858947e-05,
      "loss": 1.095,
      "step": 3031
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.13360023498535156,
      "learning_rate": 7.826900711230205e-05,
      "loss": 1.1074,
      "step": 3032
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.15417857468128204,
      "learning_rate": 7.820817981608185e-05,
      "loss": 1.0559,
      "step": 3033
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.10985743254423141,
      "learning_rate": 7.814736098355347e-05,
      "loss": 1.1968,
      "step": 3034
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.13569296896457672,
      "learning_rate": 7.808655063833832e-05,
      "loss": 1.1593,
      "step": 3035
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.15450584888458252,
      "learning_rate": 7.802574880405438e-05,
      "loss": 1.0903,
      "step": 3036
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.12837792932987213,
      "learning_rate": 7.79649555043164e-05,
      "loss": 1.1204,
      "step": 3037
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.13937190175056458,
      "learning_rate": 7.790417076273581e-05,
      "loss": 1.1431,
      "step": 3038
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1204460859298706,
      "learning_rate": 7.784339460292064e-05,
      "loss": 1.1647,
      "step": 3039
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.13571175932884216,
      "learning_rate": 7.778262704847569e-05,
      "loss": 1.1104,
      "step": 3040
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13679906725883484,
      "learning_rate": 7.772186812300244e-05,
      "loss": 1.0901,
      "step": 3041
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.12238705903291702,
      "learning_rate": 7.766111785009889e-05,
      "loss": 1.1047,
      "step": 3042
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1556052416563034,
      "learning_rate": 7.760037625335973e-05,
      "loss": 1.092,
      "step": 3043
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.11812357604503632,
      "learning_rate": 7.753964335637634e-05,
      "loss": 1.1185,
      "step": 3044
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1471358686685562,
      "learning_rate": 7.747891918273666e-05,
      "loss": 1.1871,
      "step": 3045
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.12696699798107147,
      "learning_rate": 7.741820375602523e-05,
      "loss": 1.0844,
      "step": 3046
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.15313293039798737,
      "learning_rate": 7.735749709982328e-05,
      "loss": 1.1406,
      "step": 3047
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1460648775100708,
      "learning_rate": 7.729679923770854e-05,
      "loss": 1.1344,
      "step": 3048
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13155318796634674,
      "learning_rate": 7.723611019325538e-05,
      "loss": 1.1478,
      "step": 3049
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.15642903745174408,
      "learning_rate": 7.717542999003471e-05,
      "loss": 1.101,
      "step": 3050
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13611915707588196,
      "learning_rate": 7.711475865161398e-05,
      "loss": 1.0695,
      "step": 3051
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1405603289604187,
      "learning_rate": 7.705409620155734e-05,
      "loss": 1.14,
      "step": 3052
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.141245037317276,
      "learning_rate": 7.699344266342528e-05,
      "loss": 1.1407,
      "step": 3053
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.15913034975528717,
      "learning_rate": 7.693279806077503e-05,
      "loss": 1.0994,
      "step": 3054
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.118376225233078,
      "learning_rate": 7.687216241716021e-05,
      "loss": 1.1664,
      "step": 3055
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13579928874969482,
      "learning_rate": 7.681153575613098e-05,
      "loss": 1.0316,
      "step": 3056
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.16381843388080597,
      "learning_rate": 7.675091810123404e-05,
      "loss": 1.151,
      "step": 3057
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13552571833133698,
      "learning_rate": 7.669030947601265e-05,
      "loss": 1.1136,
      "step": 3058
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13420821726322174,
      "learning_rate": 7.662970990400646e-05,
      "loss": 1.1828,
      "step": 3059
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.12594519555568695,
      "learning_rate": 7.656911940875162e-05,
      "loss": 1.1207,
      "step": 3060
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.17036749422550201,
      "learning_rate": 7.650853801378084e-05,
      "loss": 1.1028,
      "step": 3061
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.10323833674192429,
      "learning_rate": 7.644796574262321e-05,
      "loss": 1.0542,
      "step": 3062
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.10035990178585052,
      "learning_rate": 7.638740261880423e-05,
      "loss": 1.2011,
      "step": 3063
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.14981535077095032,
      "learning_rate": 7.632684866584605e-05,
      "loss": 1.1397,
      "step": 3064
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13304726779460907,
      "learning_rate": 7.626630390726703e-05,
      "loss": 1.0792,
      "step": 3065
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1461562067270279,
      "learning_rate": 7.620576836658212e-05,
      "loss": 1.1454,
      "step": 3066
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.15838885307312012,
      "learning_rate": 7.614524206730259e-05,
      "loss": 1.163,
      "step": 3067
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.15485598146915436,
      "learning_rate": 7.608472503293614e-05,
      "loss": 1.0623,
      "step": 3068
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.09722132980823517,
      "learning_rate": 7.60242172869869e-05,
      "loss": 1.0855,
      "step": 3069
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13597948849201202,
      "learning_rate": 7.596371885295541e-05,
      "loss": 1.1786,
      "step": 3070
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.11275961250066757,
      "learning_rate": 7.590322975433857e-05,
      "loss": 1.0979,
      "step": 3071
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.10720642656087875,
      "learning_rate": 7.584275001462961e-05,
      "loss": 1.0749,
      "step": 3072
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.11493778228759766,
      "learning_rate": 7.578227965731819e-05,
      "loss": 1.0892,
      "step": 3073
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.15408627688884735,
      "learning_rate": 7.572181870589028e-05,
      "loss": 1.0788,
      "step": 3074
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.16464486718177795,
      "learning_rate": 7.56613671838282e-05,
      "loss": 1.1198,
      "step": 3075
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.15451021492481232,
      "learning_rate": 7.560092511461068e-05,
      "loss": 1.0597,
      "step": 3076
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.0976378545165062,
      "learning_rate": 7.554049252171269e-05,
      "loss": 1.0872,
      "step": 3077
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.12313945591449738,
      "learning_rate": 7.548006942860557e-05,
      "loss": 1.0907,
      "step": 3078
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.10793454200029373,
      "learning_rate": 7.541965585875694e-05,
      "loss": 1.1486,
      "step": 3079
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.14201770722866058,
      "learning_rate": 7.535925183563073e-05,
      "loss": 1.1973,
      "step": 3080
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.12367488443851471,
      "learning_rate": 7.529885738268714e-05,
      "loss": 1.1597,
      "step": 3081
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.15119250118732452,
      "learning_rate": 7.523847252338274e-05,
      "loss": 1.1431,
      "step": 3082
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.12981334328651428,
      "learning_rate": 7.51780972811703e-05,
      "loss": 1.1453,
      "step": 3083
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.1257932484149933,
      "learning_rate": 7.511773167949884e-05,
      "loss": 1.1129,
      "step": 3084
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.12904268503189087,
      "learning_rate": 7.505737574181369e-05,
      "loss": 1.0684,
      "step": 3085
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.14171507954597473,
      "learning_rate": 7.499702949155633e-05,
      "loss": 1.1182,
      "step": 3086
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13538165390491486,
      "learning_rate": 7.493669295216467e-05,
      "loss": 1.1192,
      "step": 3087
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.11397839337587357,
      "learning_rate": 7.487636614707266e-05,
      "loss": 1.0889,
      "step": 3088
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.07297320663928986,
      "learning_rate": 7.48160490997105e-05,
      "loss": 1.0652,
      "step": 3089
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.13700374960899353,
      "learning_rate": 7.475574183350471e-05,
      "loss": 1.1421,
      "step": 3090
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.09713956713676453,
      "learning_rate": 7.469544437187789e-05,
      "loss": 1.0509,
      "step": 3091
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.12176094949245453,
      "learning_rate": 7.463515673824888e-05,
      "loss": 1.1472,
      "step": 3092
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12531732022762299,
      "learning_rate": 7.457487895603272e-05,
      "loss": 1.0957,
      "step": 3093
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12737302482128143,
      "learning_rate": 7.45146110486406e-05,
      "loss": 1.1002,
      "step": 3094
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.13299450278282166,
      "learning_rate": 7.44543530394799e-05,
      "loss": 1.1712,
      "step": 3095
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.11390195786952972,
      "learning_rate": 7.439410495195412e-05,
      "loss": 1.151,
      "step": 3096
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1376233845949173,
      "learning_rate": 7.433386680946288e-05,
      "loss": 1.1795,
      "step": 3097
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.14209550619125366,
      "learning_rate": 7.427363863540202e-05,
      "loss": 1.1505,
      "step": 3098
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.11758130043745041,
      "learning_rate": 7.421342045316351e-05,
      "loss": 1.1276,
      "step": 3099
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.09066219627857208,
      "learning_rate": 7.415321228613535e-05,
      "loss": 1.121,
      "step": 3100
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12903723120689392,
      "learning_rate": 7.409301415770167e-05,
      "loss": 1.148,
      "step": 3101
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.15833908319473267,
      "learning_rate": 7.40328260912428e-05,
      "loss": 1.1405,
      "step": 3102
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1495690643787384,
      "learning_rate": 7.397264811013506e-05,
      "loss": 1.0733,
      "step": 3103
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1345140039920807,
      "learning_rate": 7.391248023775083e-05,
      "loss": 1.1372,
      "step": 3104
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.14087404310703278,
      "learning_rate": 7.385232249745872e-05,
      "loss": 1.1504,
      "step": 3105
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12010256201028824,
      "learning_rate": 7.379217491262325e-05,
      "loss": 1.2574,
      "step": 3106
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.15402811765670776,
      "learning_rate": 7.373203750660505e-05,
      "loss": 1.1641,
      "step": 3107
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.13803042471408844,
      "learning_rate": 7.367191030276079e-05,
      "loss": 1.1592,
      "step": 3108
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1259058266878128,
      "learning_rate": 7.361179332444318e-05,
      "loss": 1.1631,
      "step": 3109
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.11853806674480438,
      "learning_rate": 7.355168659500095e-05,
      "loss": 1.1948,
      "step": 3110
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12641897797584534,
      "learning_rate": 7.349159013777891e-05,
      "loss": 1.1528,
      "step": 3111
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.13772159814834595,
      "learning_rate": 7.343150397611782e-05,
      "loss": 1.155,
      "step": 3112
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12880145013332367,
      "learning_rate": 7.337142813335444e-05,
      "loss": 1.0725,
      "step": 3113
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12287609279155731,
      "learning_rate": 7.33113626328215e-05,
      "loss": 1.1406,
      "step": 3114
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.14896930754184723,
      "learning_rate": 7.325130749784782e-05,
      "loss": 1.1403,
      "step": 3115
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.13467998802661896,
      "learning_rate": 7.319126275175801e-05,
      "loss": 1.0713,
      "step": 3116
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1375690996646881,
      "learning_rate": 7.31312284178729e-05,
      "loss": 1.1459,
      "step": 3117
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12281427532434464,
      "learning_rate": 7.307120451950901e-05,
      "loss": 1.0839,
      "step": 3118
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.14163446426391602,
      "learning_rate": 7.301119107997905e-05,
      "loss": 1.1131,
      "step": 3119
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.13287220895290375,
      "learning_rate": 7.295118812259145e-05,
      "loss": 1.017,
      "step": 3120
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.11161144077777863,
      "learning_rate": 7.289119567065068e-05,
      "loss": 1.0608,
      "step": 3121
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.16100452840328217,
      "learning_rate": 7.283121374745715e-05,
      "loss": 1.163,
      "step": 3122
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1379525065422058,
      "learning_rate": 7.277124237630712e-05,
      "loss": 1.1142,
      "step": 3123
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.13227707147598267,
      "learning_rate": 7.271128158049283e-05,
      "loss": 1.125,
      "step": 3124
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.16849654912948608,
      "learning_rate": 7.265133138330233e-05,
      "loss": 1.1585,
      "step": 3125
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.14249016344547272,
      "learning_rate": 7.259139180801955e-05,
      "loss": 1.1204,
      "step": 3126
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1181040108203888,
      "learning_rate": 7.253146287792434e-05,
      "loss": 1.1432,
      "step": 3127
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12684883177280426,
      "learning_rate": 7.247154461629247e-05,
      "loss": 1.1385,
      "step": 3128
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.16241246461868286,
      "learning_rate": 7.241163704639546e-05,
      "loss": 1.0938,
      "step": 3129
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12615494430065155,
      "learning_rate": 7.23517401915007e-05,
      "loss": 1.1157,
      "step": 3130
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.146034836769104,
      "learning_rate": 7.229185407487148e-05,
      "loss": 1.0453,
      "step": 3131
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.10992443561553955,
      "learning_rate": 7.223197871976689e-05,
      "loss": 1.1221,
      "step": 3132
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.11803296953439713,
      "learning_rate": 7.21721141494417e-05,
      "loss": 1.2071,
      "step": 3133
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12742143869400024,
      "learning_rate": 7.211226038714678e-05,
      "loss": 1.1241,
      "step": 3134
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.140217125415802,
      "learning_rate": 7.205241745612856e-05,
      "loss": 1.1607,
      "step": 3135
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.13667485117912292,
      "learning_rate": 7.199258537962936e-05,
      "loss": 1.1037,
      "step": 3136
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.11872421950101852,
      "learning_rate": 7.193276418088729e-05,
      "loss": 1.1678,
      "step": 3137
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.14319206774234772,
      "learning_rate": 7.187295388313617e-05,
      "loss": 1.1061,
      "step": 3138
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12596692144870758,
      "learning_rate": 7.181315450960562e-05,
      "loss": 1.0566,
      "step": 3139
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.13436712324619293,
      "learning_rate": 7.175336608352112e-05,
      "loss": 1.0639,
      "step": 3140
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1472463756799698,
      "learning_rate": 7.169358862810373e-05,
      "loss": 1.2092,
      "step": 3141
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.17255108058452606,
      "learning_rate": 7.163382216657034e-05,
      "loss": 1.2327,
      "step": 3142
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.11691432446241379,
      "learning_rate": 7.157406672213356e-05,
      "loss": 1.1123,
      "step": 3143
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.12767711281776428,
      "learning_rate": 7.151432231800172e-05,
      "loss": 1.2487,
      "step": 3144
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.11608356237411499,
      "learning_rate": 7.145458897737881e-05,
      "loss": 1.1522,
      "step": 3145
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.10779350996017456,
      "learning_rate": 7.139486672346466e-05,
      "loss": 1.1354,
      "step": 3146
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.11618991196155548,
      "learning_rate": 7.133515557945463e-05,
      "loss": 1.1712,
      "step": 3147
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.11943051964044571,
      "learning_rate": 7.12754555685399e-05,
      "loss": 1.127,
      "step": 3148
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13316357135772705,
      "learning_rate": 7.121576671390722e-05,
      "loss": 1.2226,
      "step": 3149
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.12842348217964172,
      "learning_rate": 7.115608903873905e-05,
      "loss": 1.0888,
      "step": 3150
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13176463544368744,
      "learning_rate": 7.109642256621353e-05,
      "loss": 1.154,
      "step": 3151
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13316403329372406,
      "learning_rate": 7.103676731950443e-05,
      "loss": 1.1547,
      "step": 3152
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.12171182781457901,
      "learning_rate": 7.097712332178117e-05,
      "loss": 1.138,
      "step": 3153
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.16075962781906128,
      "learning_rate": 7.09174905962088e-05,
      "loss": 1.12,
      "step": 3154
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.10293340682983398,
      "learning_rate": 7.085786916594794e-05,
      "loss": 1.1614,
      "step": 3155
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14467363059520721,
      "learning_rate": 7.07982590541549e-05,
      "loss": 1.1255,
      "step": 3156
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1316087245941162,
      "learning_rate": 7.073866028398153e-05,
      "loss": 1.1439,
      "step": 3157
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.12199747562408447,
      "learning_rate": 7.067907287857535e-05,
      "loss": 1.0856,
      "step": 3158
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1482880562543869,
      "learning_rate": 7.061949686107938e-05,
      "loss": 1.1608,
      "step": 3159
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14204393327236176,
      "learning_rate": 7.055993225463231e-05,
      "loss": 1.1215,
      "step": 3160
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1387970745563507,
      "learning_rate": 7.050037908236831e-05,
      "loss": 1.1689,
      "step": 3161
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.15924763679504395,
      "learning_rate": 7.04408373674171e-05,
      "loss": 1.0641,
      "step": 3162
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1471719592809677,
      "learning_rate": 7.03813071329041e-05,
      "loss": 1.1901,
      "step": 3163
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.15984654426574707,
      "learning_rate": 7.032178840195008e-05,
      "loss": 1.222,
      "step": 3164
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.12606939673423767,
      "learning_rate": 7.026228119767148e-05,
      "loss": 1.162,
      "step": 3165
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1434328556060791,
      "learning_rate": 7.020278554318023e-05,
      "loss": 1.1081,
      "step": 3166
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13255742192268372,
      "learning_rate": 7.014330146158366e-05,
      "loss": 1.1165,
      "step": 3167
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13206148147583008,
      "learning_rate": 7.008382897598477e-05,
      "loss": 1.1287,
      "step": 3168
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1100602000951767,
      "learning_rate": 7.0024368109482e-05,
      "loss": 1.1552,
      "step": 3169
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.11580537259578705,
      "learning_rate": 6.996491888516927e-05,
      "loss": 1.1819,
      "step": 3170
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.11574002355337143,
      "learning_rate": 6.990548132613592e-05,
      "loss": 1.1159,
      "step": 3171
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13514062762260437,
      "learning_rate": 6.984605545546686e-05,
      "loss": 1.1467,
      "step": 3172
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.10983733087778091,
      "learning_rate": 6.97866412962424e-05,
      "loss": 1.1724,
      "step": 3173
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14604881405830383,
      "learning_rate": 6.972723887153828e-05,
      "loss": 1.1365,
      "step": 3174
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.12264881283044815,
      "learning_rate": 6.966784820442577e-05,
      "loss": 1.0807,
      "step": 3175
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.12052877247333527,
      "learning_rate": 6.960846931797152e-05,
      "loss": 1.059,
      "step": 3176
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14383342862129211,
      "learning_rate": 6.954910223523756e-05,
      "loss": 1.1268,
      "step": 3177
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13464102149009705,
      "learning_rate": 6.948974697928143e-05,
      "loss": 1.2094,
      "step": 3178
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14376285672187805,
      "learning_rate": 6.943040357315598e-05,
      "loss": 1.2117,
      "step": 3179
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1384584903717041,
      "learning_rate": 6.937107203990952e-05,
      "loss": 1.1472,
      "step": 3180
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13048480451107025,
      "learning_rate": 6.931175240258577e-05,
      "loss": 1.1998,
      "step": 3181
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.15550027787685394,
      "learning_rate": 6.925244468422376e-05,
      "loss": 1.1225,
      "step": 3182
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.16620655357837677,
      "learning_rate": 6.919314890785793e-05,
      "loss": 1.1312,
      "step": 3183
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14561057090759277,
      "learning_rate": 6.913386509651807e-05,
      "loss": 1.1139,
      "step": 3184
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13790978491306305,
      "learning_rate": 6.907459327322934e-05,
      "loss": 1.1533,
      "step": 3185
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13717198371887207,
      "learning_rate": 6.90153334610122e-05,
      "loss": 1.1622,
      "step": 3186
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14986000955104828,
      "learning_rate": 6.895608568288254e-05,
      "loss": 1.0975,
      "step": 3187
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.127268984913826,
      "learning_rate": 6.889684996185148e-05,
      "loss": 1.1179,
      "step": 3188
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.11296997964382172,
      "learning_rate": 6.88376263209255e-05,
      "loss": 1.1401,
      "step": 3189
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14191731810569763,
      "learning_rate": 6.877841478310638e-05,
      "loss": 1.1403,
      "step": 3190
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14538687467575073,
      "learning_rate": 6.871921537139116e-05,
      "loss": 1.2255,
      "step": 3191
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.13731688261032104,
      "learning_rate": 6.866002810877225e-05,
      "loss": 1.117,
      "step": 3192
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.10774785280227661,
      "learning_rate": 6.860085301823729e-05,
      "loss": 1.0699,
      "step": 3193
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14090676605701447,
      "learning_rate": 6.854169012276923e-05,
      "loss": 1.1958,
      "step": 3194
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1746576726436615,
      "learning_rate": 6.848253944534622e-05,
      "loss": 1.1568,
      "step": 3195
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.11302357912063599,
      "learning_rate": 6.84234010089417e-05,
      "loss": 1.0999,
      "step": 3196
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12581758201122284,
      "learning_rate": 6.836427483652436e-05,
      "loss": 1.1125,
      "step": 3197
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13514432311058044,
      "learning_rate": 6.830516095105816e-05,
      "loss": 1.1341,
      "step": 3198
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.16563333570957184,
      "learning_rate": 6.824605937550223e-05,
      "loss": 1.1019,
      "step": 3199
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12999729812145233,
      "learning_rate": 6.818697013281092e-05,
      "loss": 1.1373,
      "step": 3200
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12904797494411469,
      "learning_rate": 6.812789324593386e-05,
      "loss": 1.147,
      "step": 3201
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.14100240170955658,
      "learning_rate": 6.80688287378158e-05,
      "loss": 1.1311,
      "step": 3202
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12439804524183273,
      "learning_rate": 6.800977663139666e-05,
      "loss": 1.1107,
      "step": 3203
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.15293355286121368,
      "learning_rate": 6.79507369496117e-05,
      "loss": 1.1387,
      "step": 3204
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1634746491909027,
      "learning_rate": 6.789170971539118e-05,
      "loss": 1.1428,
      "step": 3205
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1179853305220604,
      "learning_rate": 6.783269495166065e-05,
      "loss": 1.113,
      "step": 3206
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.14051033556461334,
      "learning_rate": 6.777369268134076e-05,
      "loss": 1.1078,
      "step": 3207
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.11623050272464752,
      "learning_rate": 6.771470292734723e-05,
      "loss": 1.1575,
      "step": 3208
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1294080764055252,
      "learning_rate": 6.765572571259106e-05,
      "loss": 1.1396,
      "step": 3209
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12362304329872131,
      "learning_rate": 6.759676105997834e-05,
      "loss": 1.1576,
      "step": 3210
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.09441171586513519,
      "learning_rate": 6.753780899241027e-05,
      "loss": 1.1164,
      "step": 3211
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.16736441850662231,
      "learning_rate": 6.74788695327831e-05,
      "loss": 1.0529,
      "step": 3212
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13250583410263062,
      "learning_rate": 6.741994270398826e-05,
      "loss": 1.1047,
      "step": 3213
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12962834537029266,
      "learning_rate": 6.736102852891227e-05,
      "loss": 1.1781,
      "step": 3214
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12597373127937317,
      "learning_rate": 6.730212703043665e-05,
      "loss": 1.015,
      "step": 3215
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13718105852603912,
      "learning_rate": 6.724323823143818e-05,
      "loss": 1.0794,
      "step": 3216
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.15005137026309967,
      "learning_rate": 6.718436215478848e-05,
      "loss": 1.0365,
      "step": 3217
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.16670021414756775,
      "learning_rate": 6.712549882335441e-05,
      "loss": 1.283,
      "step": 3218
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.16761347651481628,
      "learning_rate": 6.70666482599978e-05,
      "loss": 1.1268,
      "step": 3219
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1289609968662262,
      "learning_rate": 6.700781048757547e-05,
      "loss": 1.1194,
      "step": 3220
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12901844084262848,
      "learning_rate": 6.69489855289394e-05,
      "loss": 1.1174,
      "step": 3221
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12284760922193527,
      "learning_rate": 6.689017340693648e-05,
      "loss": 1.0633,
      "step": 3222
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.148489847779274,
      "learning_rate": 6.683137414440872e-05,
      "loss": 1.0686,
      "step": 3223
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.1075606644153595,
      "learning_rate": 6.677258776419305e-05,
      "loss": 1.0947,
      "step": 3224
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.16052687168121338,
      "learning_rate": 6.671381428912138e-05,
      "loss": 1.1597,
      "step": 3225
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13335396349430084,
      "learning_rate": 6.66550537420207e-05,
      "loss": 1.0885,
      "step": 3226
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13390296697616577,
      "learning_rate": 6.659630614571288e-05,
      "loss": 1.1366,
      "step": 3227
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.10447201877832413,
      "learning_rate": 6.653757152301487e-05,
      "loss": 1.2667,
      "step": 3228
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12866590917110443,
      "learning_rate": 6.647884989673849e-05,
      "loss": 1.0423,
      "step": 3229
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.15122678875923157,
      "learning_rate": 6.642014128969055e-05,
      "loss": 1.1585,
      "step": 3230
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.15397953987121582,
      "learning_rate": 6.63614457246728e-05,
      "loss": 1.1383,
      "step": 3231
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.14140911400318146,
      "learning_rate": 6.630276322448188e-05,
      "loss": 1.0959,
      "step": 3232
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12583880126476288,
      "learning_rate": 6.624409381190945e-05,
      "loss": 1.1119,
      "step": 3233
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13709194958209991,
      "learning_rate": 6.618543750974202e-05,
      "loss": 1.1163,
      "step": 3234
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12085220962762833,
      "learning_rate": 6.612679434076102e-05,
      "loss": 1.0377,
      "step": 3235
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.15166409313678741,
      "learning_rate": 6.606816432774278e-05,
      "loss": 1.1235,
      "step": 3236
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.15722958743572235,
      "learning_rate": 6.600954749345851e-05,
      "loss": 1.1351,
      "step": 3237
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.14774936437606812,
      "learning_rate": 6.595094386067429e-05,
      "loss": 1.179,
      "step": 3238
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12869133055210114,
      "learning_rate": 6.589235345215117e-05,
      "loss": 1.221,
      "step": 3239
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.09877712279558182,
      "learning_rate": 6.583377629064494e-05,
      "loss": 1.1634,
      "step": 3240
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.11171898245811462,
      "learning_rate": 6.57752123989063e-05,
      "loss": 1.1405,
      "step": 3241
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.16063664853572845,
      "learning_rate": 6.571666179968079e-05,
      "loss": 1.2087,
      "step": 3242
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.12422078102827072,
      "learning_rate": 6.56581245157088e-05,
      "loss": 1.1407,
      "step": 3243
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13524287939071655,
      "learning_rate": 6.55996005697255e-05,
      "loss": 1.091,
      "step": 3244
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.16574062407016754,
      "learning_rate": 6.554108998446095e-05,
      "loss": 1.0347,
      "step": 3245
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13731256127357483,
      "learning_rate": 6.548259278264e-05,
      "loss": 1.0509,
      "step": 3246
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.14902257919311523,
      "learning_rate": 6.542410898698226e-05,
      "loss": 1.2116,
      "step": 3247
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13027192652225494,
      "learning_rate": 6.536563862020218e-05,
      "loss": 1.1179,
      "step": 3248
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1394287794828415,
      "learning_rate": 6.530718170500895e-05,
      "loss": 1.0754,
      "step": 3249
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1549542248249054,
      "learning_rate": 6.524873826410658e-05,
      "loss": 1.1334,
      "step": 3250
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.12252933531999588,
      "learning_rate": 6.519030832019383e-05,
      "loss": 1.1748,
      "step": 3251
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.18784548342227936,
      "learning_rate": 6.513189189596423e-05,
      "loss": 1.1604,
      "step": 3252
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1235223338007927,
      "learning_rate": 6.507348901410604e-05,
      "loss": 1.138,
      "step": 3253
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.11458469182252884,
      "learning_rate": 6.501509969730224e-05,
      "loss": 1.1543,
      "step": 3254
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1677645444869995,
      "learning_rate": 6.49567239682306e-05,
      "loss": 1.0309,
      "step": 3255
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.12833596765995026,
      "learning_rate": 6.489836184956352e-05,
      "loss": 1.0981,
      "step": 3256
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.13872838020324707,
      "learning_rate": 6.484001336396828e-05,
      "loss": 1.1225,
      "step": 3257
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1556912511587143,
      "learning_rate": 6.478167853410668e-05,
      "loss": 1.1546,
      "step": 3258
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.12202286720275879,
      "learning_rate": 6.472335738263534e-05,
      "loss": 1.0847,
      "step": 3259
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1335546374320984,
      "learning_rate": 6.466504993220548e-05,
      "loss": 1.0525,
      "step": 3260
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.12513428926467896,
      "learning_rate": 6.460675620546305e-05,
      "loss": 1.1277,
      "step": 3261
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1202024519443512,
      "learning_rate": 6.454847622504867e-05,
      "loss": 1.1046,
      "step": 3262
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.12865856289863586,
      "learning_rate": 6.449021001359763e-05,
      "loss": 1.1426,
      "step": 3263
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.14883753657341003,
      "learning_rate": 6.443195759373984e-05,
      "loss": 1.155,
      "step": 3264
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.14921967685222626,
      "learning_rate": 6.43737189880999e-05,
      "loss": 1.2224,
      "step": 3265
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.10083440691232681,
      "learning_rate": 6.431549421929694e-05,
      "loss": 1.1386,
      "step": 3266
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.13920369744300842,
      "learning_rate": 6.42572833099448e-05,
      "loss": 1.124,
      "step": 3267
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.14771795272827148,
      "learning_rate": 6.419908628265203e-05,
      "loss": 1.1595,
      "step": 3268
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.11332732439041138,
      "learning_rate": 6.41409031600216e-05,
      "loss": 1.0621,
      "step": 3269
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.11158725619316101,
      "learning_rate": 6.408273396465117e-05,
      "loss": 1.0581,
      "step": 3270
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.08753504604101181,
      "learning_rate": 6.4024578719133e-05,
      "loss": 1.0673,
      "step": 3271
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1486949771642685,
      "learning_rate": 6.396643744605391e-05,
      "loss": 1.1308,
      "step": 3272
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1397649198770523,
      "learning_rate": 6.390831016799526e-05,
      "loss": 1.173,
      "step": 3273
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.13930174708366394,
      "learning_rate": 6.38501969075331e-05,
      "loss": 1.1236,
      "step": 3274
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1296149641275406,
      "learning_rate": 6.379209768723791e-05,
      "loss": 1.1483,
      "step": 3275
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.07698202133178711,
      "learning_rate": 6.373401252967475e-05,
      "loss": 1.143,
      "step": 3276
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.17480066418647766,
      "learning_rate": 6.367594145740324e-05,
      "loss": 1.2073,
      "step": 3277
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1601085215806961,
      "learning_rate": 6.361788449297747e-05,
      "loss": 1.1444,
      "step": 3278
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.16382262110710144,
      "learning_rate": 6.355984165894613e-05,
      "loss": 1.0932,
      "step": 3279
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.13617300987243652,
      "learning_rate": 6.350181297785241e-05,
      "loss": 1.1011,
      "step": 3280
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1575622409582138,
      "learning_rate": 6.344379847223398e-05,
      "loss": 1.0475,
      "step": 3281
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.11010781675577164,
      "learning_rate": 6.338579816462298e-05,
      "loss": 1.1222,
      "step": 3282
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.127905011177063,
      "learning_rate": 6.332781207754605e-05,
      "loss": 1.1914,
      "step": 3283
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1224808394908905,
      "learning_rate": 6.326984023352435e-05,
      "loss": 1.0979,
      "step": 3284
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.15700477361679077,
      "learning_rate": 6.321188265507342e-05,
      "loss": 1.1509,
      "step": 3285
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.14495186507701874,
      "learning_rate": 6.31539393647034e-05,
      "loss": 1.0803,
      "step": 3286
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.15799562633037567,
      "learning_rate": 6.309601038491874e-05,
      "loss": 1.147,
      "step": 3287
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.09922036528587341,
      "learning_rate": 6.303809573821842e-05,
      "loss": 1.2058,
      "step": 3288
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.16903764009475708,
      "learning_rate": 6.298019544709579e-05,
      "loss": 1.1135,
      "step": 3289
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.17682762444019318,
      "learning_rate": 6.292230953403866e-05,
      "loss": 1.0786,
      "step": 3290
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.12633714079856873,
      "learning_rate": 6.286443802152926e-05,
      "loss": 1.1208,
      "step": 3291
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.16338679194450378,
      "learning_rate": 6.280658093204422e-05,
      "loss": 1.109,
      "step": 3292
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.16803736984729767,
      "learning_rate": 6.274873828805458e-05,
      "loss": 1.1353,
      "step": 3293
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.14700336754322052,
      "learning_rate": 6.269091011202575e-05,
      "loss": 1.1334,
      "step": 3294
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.11886899173259735,
      "learning_rate": 6.26330964264175e-05,
      "loss": 1.0413,
      "step": 3295
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.1240144670009613,
      "learning_rate": 6.257529725368405e-05,
      "loss": 1.1292,
      "step": 3296
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.12482820451259613,
      "learning_rate": 6.251751261627385e-05,
      "loss": 1.1563,
      "step": 3297
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.12348951399326324,
      "learning_rate": 6.245974253662987e-05,
      "loss": 1.0749,
      "step": 3298
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.10131867229938507,
      "learning_rate": 6.240198703718931e-05,
      "loss": 1.0126,
      "step": 3299
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.13314135372638702,
      "learning_rate": 6.234424614038375e-05,
      "loss": 1.0527,
      "step": 3300
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.13976728916168213,
      "learning_rate": 6.22865198686391e-05,
      "loss": 1.0919,
      "step": 3301
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.13264711201190948,
      "learning_rate": 6.22288082443755e-05,
      "loss": 1.1873,
      "step": 3302
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.13108648359775543,
      "learning_rate": 6.217111129000759e-05,
      "loss": 1.0429,
      "step": 3303
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1253369152545929,
      "learning_rate": 6.211342902794413e-05,
      "loss": 1.1256,
      "step": 3304
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12806548178195953,
      "learning_rate": 6.205576148058828e-05,
      "loss": 1.1222,
      "step": 3305
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.10991417616605759,
      "learning_rate": 6.199810867033746e-05,
      "loss": 1.1249,
      "step": 3306
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.13168686628341675,
      "learning_rate": 6.19404706195833e-05,
      "loss": 1.1194,
      "step": 3307
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.11135033518075943,
      "learning_rate": 6.188284735071177e-05,
      "loss": 1.106,
      "step": 3308
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.11596009880304337,
      "learning_rate": 6.182523888610316e-05,
      "loss": 1.3289,
      "step": 3309
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1270655393600464,
      "learning_rate": 6.176764524813187e-05,
      "loss": 1.2187,
      "step": 3310
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12120752781629562,
      "learning_rate": 6.171006645916661e-05,
      "loss": 1.1522,
      "step": 3311
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.13903608918190002,
      "learning_rate": 6.165250254157031e-05,
      "loss": 1.1452,
      "step": 3312
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.11722181737422943,
      "learning_rate": 6.159495351770017e-05,
      "loss": 1.1823,
      "step": 3313
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12782073020935059,
      "learning_rate": 6.15374194099075e-05,
      "loss": 1.0774,
      "step": 3314
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.10886353999376297,
      "learning_rate": 6.147990024053796e-05,
      "loss": 1.0648,
      "step": 3315
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.15017598867416382,
      "learning_rate": 6.142239603193127e-05,
      "loss": 1.1986,
      "step": 3316
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1390228271484375,
      "learning_rate": 6.136490680642146e-05,
      "loss": 1.1466,
      "step": 3317
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12900927662849426,
      "learning_rate": 6.130743258633667e-05,
      "loss": 1.1736,
      "step": 3318
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.11640990525484085,
      "learning_rate": 6.124997339399916e-05,
      "loss": 1.1422,
      "step": 3319
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.13949060440063477,
      "learning_rate": 6.11925292517255e-05,
      "loss": 1.077,
      "step": 3320
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.14943422377109528,
      "learning_rate": 6.113510018182627e-05,
      "loss": 1.1872,
      "step": 3321
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12605266273021698,
      "learning_rate": 6.107768620660632e-05,
      "loss": 1.1614,
      "step": 3322
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.15124519169330597,
      "learning_rate": 6.102028734836456e-05,
      "loss": 1.1259,
      "step": 3323
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1389375627040863,
      "learning_rate": 6.0962903629394e-05,
      "loss": 1.0479,
      "step": 3324
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.14357422292232513,
      "learning_rate": 6.090553507198187e-05,
      "loss": 1.1066,
      "step": 3325
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.11105064302682877,
      "learning_rate": 6.084818169840938e-05,
      "loss": 1.1578,
      "step": 3326
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12999515235424042,
      "learning_rate": 6.079084353095201e-05,
      "loss": 1.1188,
      "step": 3327
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.13105817139148712,
      "learning_rate": 6.07335205918792e-05,
      "loss": 1.1407,
      "step": 3328
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.14291362464427948,
      "learning_rate": 6.0676212903454543e-05,
      "loss": 1.1255,
      "step": 3329
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.15723364055156708,
      "learning_rate": 6.061892048793567e-05,
      "loss": 1.1215,
      "step": 3330
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.15037134289741516,
      "learning_rate": 6.056164336757426e-05,
      "loss": 1.1526,
      "step": 3331
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.11242007464170456,
      "learning_rate": 6.050438156461613e-05,
      "loss": 1.1936,
      "step": 3332
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1558246910572052,
      "learning_rate": 6.0447135101301077e-05,
      "loss": 1.109,
      "step": 3333
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.15401208400726318,
      "learning_rate": 6.038990399986302e-05,
      "loss": 1.0785,
      "step": 3334
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.14298488199710846,
      "learning_rate": 6.0332688282529804e-05,
      "loss": 1.1975,
      "step": 3335
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12421760708093643,
      "learning_rate": 6.027548797152336e-05,
      "loss": 1.1642,
      "step": 3336
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.14333778619766235,
      "learning_rate": 6.0218303089059626e-05,
      "loss": 1.0404,
      "step": 3337
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12342990934848785,
      "learning_rate": 6.016113365734861e-05,
      "loss": 1.1663,
      "step": 3338
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.14015589654445648,
      "learning_rate": 6.010397969859421e-05,
      "loss": 1.1484,
      "step": 3339
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1382535994052887,
      "learning_rate": 6.004684123499436e-05,
      "loss": 1.1012,
      "step": 3340
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1385858952999115,
      "learning_rate": 5.998971828874102e-05,
      "loss": 1.1118,
      "step": 3341
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.11747649312019348,
      "learning_rate": 5.9932610882020046e-05,
      "loss": 1.0945,
      "step": 3342
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1367582231760025,
      "learning_rate": 5.987551903701127e-05,
      "loss": 1.1232,
      "step": 3343
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12851808965206146,
      "learning_rate": 5.9818442775888595e-05,
      "loss": 1.1026,
      "step": 3344
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.11883413791656494,
      "learning_rate": 5.97613821208197e-05,
      "loss": 1.1533,
      "step": 3345
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1129518374800682,
      "learning_rate": 5.9704337093966344e-05,
      "loss": 1.138,
      "step": 3346
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.13888297975063324,
      "learning_rate": 5.9647307717484143e-05,
      "loss": 1.196,
      "step": 3347
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1402166336774826,
      "learning_rate": 5.959029401352262e-05,
      "loss": 1.0742,
      "step": 3348
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.11493659764528275,
      "learning_rate": 5.9533296004225235e-05,
      "loss": 1.0548,
      "step": 3349
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1308387815952301,
      "learning_rate": 5.947631371172942e-05,
      "loss": 1.0999,
      "step": 3350
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.1535419225692749,
      "learning_rate": 5.941934715816642e-05,
      "loss": 1.1292,
      "step": 3351
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.18441419303417206,
      "learning_rate": 5.936239636566137e-05,
      "loss": 1.135,
      "step": 3352
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11359477788209915,
      "learning_rate": 5.930546135633327e-05,
      "loss": 1.1256,
      "step": 3353
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.12982383370399475,
      "learning_rate": 5.924854215229508e-05,
      "loss": 1.1109,
      "step": 3354
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.12821049988269806,
      "learning_rate": 5.91916387756535e-05,
      "loss": 1.082,
      "step": 3355
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11519841849803925,
      "learning_rate": 5.9134751248509236e-05,
      "loss": 1.1308,
      "step": 3356
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.13937592506408691,
      "learning_rate": 5.9077879592956675e-05,
      "loss": 1.0871,
      "step": 3357
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.14127492904663086,
      "learning_rate": 5.902102383108414e-05,
      "loss": 1.1256,
      "step": 3358
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.10861258208751678,
      "learning_rate": 5.896418398497377e-05,
      "loss": 1.0947,
      "step": 3359
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.17127160727977753,
      "learning_rate": 5.890736007670144e-05,
      "loss": 1.1677,
      "step": 3360
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11484725028276443,
      "learning_rate": 5.8850552128336954e-05,
      "loss": 1.1465,
      "step": 3361
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.15338543057441711,
      "learning_rate": 5.879376016194387e-05,
      "loss": 1.0852,
      "step": 3362
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.15153995156288147,
      "learning_rate": 5.873698419957952e-05,
      "loss": 1.1336,
      "step": 3363
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.15006105601787567,
      "learning_rate": 5.868022426329505e-05,
      "loss": 1.0925,
      "step": 3364
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.14359697699546814,
      "learning_rate": 5.862348037513532e-05,
      "loss": 1.2631,
      "step": 3365
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.10283967107534409,
      "learning_rate": 5.856675255713905e-05,
      "loss": 1.1972,
      "step": 3366
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1393507570028305,
      "learning_rate": 5.851004083133862e-05,
      "loss": 1.1071,
      "step": 3367
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.13581718504428864,
      "learning_rate": 5.845334521976027e-05,
      "loss": 1.1539,
      "step": 3368
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.17360487580299377,
      "learning_rate": 5.8396665744423885e-05,
      "loss": 1.1095,
      "step": 3369
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1419687718153,
      "learning_rate": 5.8340002427343164e-05,
      "loss": 1.1181,
      "step": 3370
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.10537067800760269,
      "learning_rate": 5.828335529052541e-05,
      "loss": 1.0552,
      "step": 3371
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.14028002321720123,
      "learning_rate": 5.822672435597172e-05,
      "loss": 1.1369,
      "step": 3372
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1422571837902069,
      "learning_rate": 5.817010964567702e-05,
      "loss": 1.1893,
      "step": 3373
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.12129679322242737,
      "learning_rate": 5.811351118162969e-05,
      "loss": 1.1513,
      "step": 3374
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1613542139530182,
      "learning_rate": 5.8056928985811963e-05,
      "loss": 1.1362,
      "step": 3375
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1508883833885193,
      "learning_rate": 5.8000363080199736e-05,
      "loss": 1.0735,
      "step": 3376
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.13249103724956512,
      "learning_rate": 5.79438134867625e-05,
      "loss": 1.1053,
      "step": 3377
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.12599122524261475,
      "learning_rate": 5.7887280227463484e-05,
      "loss": 1.1734,
      "step": 3378
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11785290390253067,
      "learning_rate": 5.783076332425956e-05,
      "loss": 1.1449,
      "step": 3379
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.09623392671346664,
      "learning_rate": 5.777426279910125e-05,
      "loss": 1.2148,
      "step": 3380
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1086331158876419,
      "learning_rate": 5.771777867393274e-05,
      "loss": 1.0706,
      "step": 3381
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11691765487194061,
      "learning_rate": 5.766131097069174e-05,
      "loss": 1.1268,
      "step": 3382
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1363677680492401,
      "learning_rate": 5.760485971130969e-05,
      "loss": 1.0826,
      "step": 3383
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.10677909106016159,
      "learning_rate": 5.75484249177116e-05,
      "loss": 1.1519,
      "step": 3384
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11877408623695374,
      "learning_rate": 5.749200661181611e-05,
      "loss": 1.1885,
      "step": 3385
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11342158168554306,
      "learning_rate": 5.743560481553547e-05,
      "loss": 1.1015,
      "step": 3386
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.10634198784828186,
      "learning_rate": 5.737921955077541e-05,
      "loss": 1.0842,
      "step": 3387
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.15981483459472656,
      "learning_rate": 5.732285083943536e-05,
      "loss": 1.0778,
      "step": 3388
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.14529551565647125,
      "learning_rate": 5.7266498703408325e-05,
      "loss": 1.1279,
      "step": 3389
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11509866267442703,
      "learning_rate": 5.721016316458068e-05,
      "loss": 1.1105,
      "step": 3390
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.16417481005191803,
      "learning_rate": 5.715384424483268e-05,
      "loss": 1.0845,
      "step": 3391
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.15041901171207428,
      "learning_rate": 5.7097541966037816e-05,
      "loss": 1.1072,
      "step": 3392
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.14618030190467834,
      "learning_rate": 5.7041256350063286e-05,
      "loss": 1.1309,
      "step": 3393
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.15526007115840912,
      "learning_rate": 5.698498741876982e-05,
      "loss": 1.0798,
      "step": 3394
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11102015525102615,
      "learning_rate": 5.692873519401154e-05,
      "loss": 1.1319,
      "step": 3395
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.13792793452739716,
      "learning_rate": 5.6872499697636195e-05,
      "loss": 1.1589,
      "step": 3396
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.14473731815814972,
      "learning_rate": 5.681628095148501e-05,
      "loss": 1.1091,
      "step": 3397
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.13438469171524048,
      "learning_rate": 5.6760078977392706e-05,
      "loss": 1.1355,
      "step": 3398
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11579878628253937,
      "learning_rate": 5.67038937971875e-05,
      "loss": 1.1177,
      "step": 3399
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.13246332108974457,
      "learning_rate": 5.664772543269101e-05,
      "loss": 1.1704,
      "step": 3400
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.12296942621469498,
      "learning_rate": 5.659157390571842e-05,
      "loss": 1.157,
      "step": 3401
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1583949327468872,
      "learning_rate": 5.653543923807832e-05,
      "loss": 1.1626,
      "step": 3402
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1349952667951584,
      "learning_rate": 5.6479321451572784e-05,
      "loss": 1.1381,
      "step": 3403
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.11144116520881653,
      "learning_rate": 5.6423220567997316e-05,
      "loss": 1.1108,
      "step": 3404
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.14001792669296265,
      "learning_rate": 5.636713660914087e-05,
      "loss": 1.0891,
      "step": 3405
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.12848447263240814,
      "learning_rate": 5.631106959678575e-05,
      "loss": 1.1028,
      "step": 3406
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1260291188955307,
      "learning_rate": 5.625501955270776e-05,
      "loss": 1.047,
      "step": 3407
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1477809101343155,
      "learning_rate": 5.619898649867612e-05,
      "loss": 1.0947,
      "step": 3408
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1437091827392578,
      "learning_rate": 5.614297045645339e-05,
      "loss": 1.0552,
      "step": 3409
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.12775319814682007,
      "learning_rate": 5.6086971447795624e-05,
      "loss": 1.154,
      "step": 3410
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.15196897089481354,
      "learning_rate": 5.603098949445209e-05,
      "loss": 1.1051,
      "step": 3411
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.14690417051315308,
      "learning_rate": 5.597502461816557e-05,
      "loss": 1.0446,
      "step": 3412
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1466226726770401,
      "learning_rate": 5.591907684067221e-05,
      "loss": 1.1606,
      "step": 3413
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.11835279315710068,
      "learning_rate": 5.586314618370145e-05,
      "loss": 1.1281,
      "step": 3414
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.11300622671842575,
      "learning_rate": 5.580723266897616e-05,
      "loss": 1.0368,
      "step": 3415
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.15205559134483337,
      "learning_rate": 5.575133631821243e-05,
      "loss": 1.0531,
      "step": 3416
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.13773401081562042,
      "learning_rate": 5.5695457153119804e-05,
      "loss": 1.2455,
      "step": 3417
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.14325907826423645,
      "learning_rate": 5.563959519540114e-05,
      "loss": 1.1968,
      "step": 3418
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.14824026823043823,
      "learning_rate": 5.5583750466752435e-05,
      "loss": 1.1305,
      "step": 3419
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1376131922006607,
      "learning_rate": 5.552792298886334e-05,
      "loss": 1.201,
      "step": 3420
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.11925968527793884,
      "learning_rate": 5.547211278341646e-05,
      "loss": 1.1387,
      "step": 3421
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1508665233850479,
      "learning_rate": 5.541631987208788e-05,
      "loss": 1.1335,
      "step": 3422
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.12587517499923706,
      "learning_rate": 5.536054427654698e-05,
      "loss": 1.0649,
      "step": 3423
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.11640158295631409,
      "learning_rate": 5.530478601845623e-05,
      "loss": 1.1211,
      "step": 3424
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1301594078540802,
      "learning_rate": 5.52490451194716e-05,
      "loss": 1.2971,
      "step": 3425
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1446947455406189,
      "learning_rate": 5.5193321601242156e-05,
      "loss": 1.0853,
      "step": 3426
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.13871178030967712,
      "learning_rate": 5.513761548541031e-05,
      "loss": 1.1573,
      "step": 3427
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.13991224765777588,
      "learning_rate": 5.5081926793611694e-05,
      "loss": 1.1035,
      "step": 3428
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.13684667646884918,
      "learning_rate": 5.502625554747508e-05,
      "loss": 1.08,
      "step": 3429
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.14835037291049957,
      "learning_rate": 5.497060176862259e-05,
      "loss": 1.0786,
      "step": 3430
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.12682050466537476,
      "learning_rate": 5.4914965478669475e-05,
      "loss": 1.1654,
      "step": 3431
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.15536414086818695,
      "learning_rate": 5.485934669922428e-05,
      "loss": 1.131,
      "step": 3432
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.12542201578617096,
      "learning_rate": 5.480374545188866e-05,
      "loss": 1.148,
      "step": 3433
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.12443666905164719,
      "learning_rate": 5.4748161758257544e-05,
      "loss": 1.1659,
      "step": 3434
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.14529480040073395,
      "learning_rate": 5.469259563991893e-05,
      "loss": 1.1166,
      "step": 3435
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.15246987342834473,
      "learning_rate": 5.4637047118454096e-05,
      "loss": 1.131,
      "step": 3436
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.10657120496034622,
      "learning_rate": 5.458151621543743e-05,
      "loss": 1.1181,
      "step": 3437
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1197897344827652,
      "learning_rate": 5.4526002952436526e-05,
      "loss": 1.0611,
      "step": 3438
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.16801434755325317,
      "learning_rate": 5.447050735101211e-05,
      "loss": 1.0634,
      "step": 3439
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1402294933795929,
      "learning_rate": 5.441502943271797e-05,
      "loss": 1.1574,
      "step": 3440
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.16292433440685272,
      "learning_rate": 5.4359569219101114e-05,
      "loss": 1.1486,
      "step": 3441
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1063527911901474,
      "learning_rate": 5.4304126731701665e-05,
      "loss": 1.1965,
      "step": 3442
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1718684583902359,
      "learning_rate": 5.424870199205283e-05,
      "loss": 1.1434,
      "step": 3443
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1641063094139099,
      "learning_rate": 5.4193295021681e-05,
      "loss": 1.0833,
      "step": 3444
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.1549408733844757,
      "learning_rate": 5.413790584210551e-05,
      "loss": 1.1305,
      "step": 3445
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.10325906425714493,
      "learning_rate": 5.408253447483892e-05,
      "loss": 1.2326,
      "step": 3446
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.13187433779239655,
      "learning_rate": 5.4027180941386877e-05,
      "loss": 1.0975,
      "step": 3447
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.12432653456926346,
      "learning_rate": 5.397184526324792e-05,
      "loss": 1.0967,
      "step": 3448
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.11875515431165695,
      "learning_rate": 5.391652746191398e-05,
      "loss": 1.1578,
      "step": 3449
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.11043474823236465,
      "learning_rate": 5.3861227558869695e-05,
      "loss": 1.0669,
      "step": 3450
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.15160499513149261,
      "learning_rate": 5.3805945575592975e-05,
      "loss": 1.1311,
      "step": 3451
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.11170517653226852,
      "learning_rate": 5.3750681533554735e-05,
      "loss": 1.1412,
      "step": 3452
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.12025666981935501,
      "learning_rate": 5.369543545421883e-05,
      "loss": 1.1113,
      "step": 3453
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.11796276271343231,
      "learning_rate": 5.3640207359042224e-05,
      "loss": 1.0943,
      "step": 3454
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.208832785487175,
      "learning_rate": 5.358499726947488e-05,
      "loss": 1.118,
      "step": 3455
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.11838402599096298,
      "learning_rate": 5.352980520695974e-05,
      "loss": 1.0356,
      "step": 3456
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.13626421988010406,
      "learning_rate": 5.347463119293283e-05,
      "loss": 1.1007,
      "step": 3457
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.12473402172327042,
      "learning_rate": 5.3419475248823014e-05,
      "loss": 1.1228,
      "step": 3458
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15722514688968658,
      "learning_rate": 5.3364337396052265e-05,
      "loss": 1.0864,
      "step": 3459
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15389660000801086,
      "learning_rate": 5.3309217656035496e-05,
      "loss": 1.012,
      "step": 3460
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.10990332067012787,
      "learning_rate": 5.3254116050180555e-05,
      "loss": 1.153,
      "step": 3461
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.14037910103797913,
      "learning_rate": 5.31990325998883e-05,
      "loss": 1.1788,
      "step": 3462
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15533365309238434,
      "learning_rate": 5.314396732655253e-05,
      "loss": 1.1341,
      "step": 3463
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.14236487448215485,
      "learning_rate": 5.3088920251559895e-05,
      "loss": 1.1099,
      "step": 3464
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15468370914459229,
      "learning_rate": 5.303389139629007e-05,
      "loss": 1.0497,
      "step": 3465
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15384429693222046,
      "learning_rate": 5.297888078211564e-05,
      "loss": 1.1533,
      "step": 3466
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.13327853381633759,
      "learning_rate": 5.292388843040208e-05,
      "loss": 1.0313,
      "step": 3467
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.11875534057617188,
      "learning_rate": 5.286891436250785e-05,
      "loss": 1.0645,
      "step": 3468
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.09739318490028381,
      "learning_rate": 5.281395859978414e-05,
      "loss": 1.1436,
      "step": 3469
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.13640880584716797,
      "learning_rate": 5.2759021163575186e-05,
      "loss": 1.1126,
      "step": 3470
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1227947399020195,
      "learning_rate": 5.27041020752181e-05,
      "loss": 1.1296,
      "step": 3471
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15593203902244568,
      "learning_rate": 5.2649201356042696e-05,
      "loss": 1.0634,
      "step": 3472
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.11009617149829865,
      "learning_rate": 5.259431902737194e-05,
      "loss": 1.095,
      "step": 3473
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.14053480327129364,
      "learning_rate": 5.2539455110521385e-05,
      "loss": 1.0585,
      "step": 3474
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15435515344142914,
      "learning_rate": 5.248460962679958e-05,
      "loss": 1.191,
      "step": 3475
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.16726626455783844,
      "learning_rate": 5.2429782597507905e-05,
      "loss": 1.0749,
      "step": 3476
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.11753199249505997,
      "learning_rate": 5.237497404394044e-05,
      "loss": 1.0568,
      "step": 3477
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.11749201267957687,
      "learning_rate": 5.232018398738435e-05,
      "loss": 1.0623,
      "step": 3478
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15815013647079468,
      "learning_rate": 5.2265412449119355e-05,
      "loss": 1.0826,
      "step": 3479
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.13754703104496002,
      "learning_rate": 5.221065945041811e-05,
      "loss": 1.1782,
      "step": 3480
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1932927668094635,
      "learning_rate": 5.215592501254609e-05,
      "loss": 1.1466,
      "step": 3481
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15095946192741394,
      "learning_rate": 5.2101209156761465e-05,
      "loss": 1.0621,
      "step": 3482
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.12983322143554688,
      "learning_rate": 5.2046511904315265e-05,
      "loss": 1.1366,
      "step": 3483
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.12629581987857819,
      "learning_rate": 5.1991833276451275e-05,
      "loss": 1.0759,
      "step": 3484
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.18528619408607483,
      "learning_rate": 5.193717329440604e-05,
      "loss": 1.0985,
      "step": 3485
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1254042536020279,
      "learning_rate": 5.188253197940889e-05,
      "loss": 1.0936,
      "step": 3486
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1308523267507553,
      "learning_rate": 5.182790935268185e-05,
      "loss": 1.1468,
      "step": 3487
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15475909411907196,
      "learning_rate": 5.177330543543971e-05,
      "loss": 1.122,
      "step": 3488
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.11680535227060318,
      "learning_rate": 5.171872024889004e-05,
      "loss": 1.111,
      "step": 3489
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.13510271906852722,
      "learning_rate": 5.1664153814233064e-05,
      "loss": 1.0922,
      "step": 3490
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.16770519316196442,
      "learning_rate": 5.160960615266178e-05,
      "loss": 1.1971,
      "step": 3491
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.14184562861919403,
      "learning_rate": 5.155507728536191e-05,
      "loss": 1.1503,
      "step": 3492
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15003502368927002,
      "learning_rate": 5.150056723351173e-05,
      "loss": 1.1443,
      "step": 3493
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.11301054060459137,
      "learning_rate": 5.1446076018282396e-05,
      "loss": 1.0925,
      "step": 3494
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.143320694565773,
      "learning_rate": 5.139160366083765e-05,
      "loss": 1.1897,
      "step": 3495
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.12336084246635437,
      "learning_rate": 5.133715018233393e-05,
      "loss": 1.1645,
      "step": 3496
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.14380942285060883,
      "learning_rate": 5.1282715603920374e-05,
      "loss": 1.0905,
      "step": 3497
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1260506510734558,
      "learning_rate": 5.1228299946738655e-05,
      "loss": 1.0988,
      "step": 3498
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1407264620065689,
      "learning_rate": 5.117390323192326e-05,
      "loss": 1.1488,
      "step": 3499
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1340273916721344,
      "learning_rate": 5.111952548060126e-05,
      "loss": 1.0323,
      "step": 3500
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.11854438483715057,
      "learning_rate": 5.106516671389223e-05,
      "loss": 1.1279,
      "step": 3501
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.14939025044441223,
      "learning_rate": 5.101082695290865e-05,
      "loss": 1.085,
      "step": 3502
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15693753957748413,
      "learning_rate": 5.095650621875534e-05,
      "loss": 1.057,
      "step": 3503
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.15308475494384766,
      "learning_rate": 5.09022045325299e-05,
      "loss": 1.1438,
      "step": 3504
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.13326367735862732,
      "learning_rate": 5.084792191532248e-05,
      "loss": 1.0988,
      "step": 3505
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.11390896886587143,
      "learning_rate": 5.079365838821579e-05,
      "loss": 1.1397,
      "step": 3506
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.12781569361686707,
      "learning_rate": 5.073941397228518e-05,
      "loss": 1.1096,
      "step": 3507
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.13492754101753235,
      "learning_rate": 5.068518868859854e-05,
      "loss": 1.2032,
      "step": 3508
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.09545949846506119,
      "learning_rate": 5.0630982558216363e-05,
      "loss": 1.0737,
      "step": 3509
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.14441916346549988,
      "learning_rate": 5.057679560219173e-05,
      "loss": 1.1003,
      "step": 3510
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.11824554204940796,
      "learning_rate": 5.052262784157014e-05,
      "loss": 1.163,
      "step": 3511
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2091289758682251,
      "learning_rate": 5.046847929738971e-05,
      "loss": 1.13,
      "step": 3512
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1365305483341217,
      "learning_rate": 5.041434999068126e-05,
      "loss": 1.2462,
      "step": 3513
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12036214768886566,
      "learning_rate": 5.036023994246787e-05,
      "loss": 1.1171,
      "step": 3514
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12902233004570007,
      "learning_rate": 5.0306149173765316e-05,
      "loss": 1.0851,
      "step": 3515
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.15075312554836273,
      "learning_rate": 5.0252077705581755e-05,
      "loss": 1.1211,
      "step": 3516
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.15256419777870178,
      "learning_rate": 5.019802555891798e-05,
      "loss": 1.1524,
      "step": 3517
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.14140504598617554,
      "learning_rate": 5.014399275476721e-05,
      "loss": 1.1356,
      "step": 3518
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.16081976890563965,
      "learning_rate": 5.008997931411517e-05,
      "loss": 1.1287,
      "step": 3519
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.14716291427612305,
      "learning_rate": 5.0035985257940024e-05,
      "loss": 1.1892,
      "step": 3520
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1394110471010208,
      "learning_rate": 4.998201060721253e-05,
      "loss": 1.1405,
      "step": 3521
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12303124368190765,
      "learning_rate": 4.992805538289571e-05,
      "loss": 1.1444,
      "step": 3522
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.13852068781852722,
      "learning_rate": 4.98741196059452e-05,
      "loss": 1.0706,
      "step": 3523
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.11805427074432373,
      "learning_rate": 4.982020329730904e-05,
      "loss": 1.1196,
      "step": 3524
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.11319857835769653,
      "learning_rate": 4.97663064779277e-05,
      "loss": 1.0627,
      "step": 3525
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12053787708282471,
      "learning_rate": 4.971242916873412e-05,
      "loss": 1.1152,
      "step": 3526
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.18891651928424835,
      "learning_rate": 4.965857139065354e-05,
      "loss": 1.03,
      "step": 3527
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12623463571071625,
      "learning_rate": 4.960473316460376e-05,
      "loss": 1.1192,
      "step": 3528
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12007002532482147,
      "learning_rate": 4.955091451149495e-05,
      "loss": 1.1309,
      "step": 3529
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.11422275006771088,
      "learning_rate": 4.9497115452229534e-05,
      "loss": 1.0929,
      "step": 3530
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.14968758821487427,
      "learning_rate": 4.944333600770261e-05,
      "loss": 1.1674,
      "step": 3531
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1296890377998352,
      "learning_rate": 4.938957619880138e-05,
      "loss": 1.1079,
      "step": 3532
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1114138588309288,
      "learning_rate": 4.933583604640557e-05,
      "loss": 1.0932,
      "step": 3533
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.131681427359581,
      "learning_rate": 4.928211557138728e-05,
      "loss": 1.0901,
      "step": 3534
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.13624626398086548,
      "learning_rate": 4.922841479461083e-05,
      "loss": 1.1675,
      "step": 3535
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.13296513259410858,
      "learning_rate": 4.917473373693304e-05,
      "loss": 1.0944,
      "step": 3536
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.13851214945316315,
      "learning_rate": 4.912107241920302e-05,
      "loss": 1.1437,
      "step": 3537
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1485777646303177,
      "learning_rate": 4.906743086226218e-05,
      "loss": 1.1286,
      "step": 3538
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.15606604516506195,
      "learning_rate": 4.901380908694434e-05,
      "loss": 1.1072,
      "step": 3539
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.11277315020561218,
      "learning_rate": 4.8960207114075495e-05,
      "loss": 1.1527,
      "step": 3540
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.127786323428154,
      "learning_rate": 4.890662496447407e-05,
      "loss": 1.2444,
      "step": 3541
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.14363984763622284,
      "learning_rate": 4.885306265895076e-05,
      "loss": 1.1034,
      "step": 3542
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.15496820211410522,
      "learning_rate": 4.879952021830856e-05,
      "loss": 1.1199,
      "step": 3543
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.13209088146686554,
      "learning_rate": 4.8745997663342755e-05,
      "loss": 1.144,
      "step": 3544
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.19076742231845856,
      "learning_rate": 4.8692495014840825e-05,
      "loss": 1.113,
      "step": 3545
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12979991734027863,
      "learning_rate": 4.863901229358261e-05,
      "loss": 1.0957,
      "step": 3546
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12904657423496246,
      "learning_rate": 4.858554952034019e-05,
      "loss": 1.1722,
      "step": 3547
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.11486652493476868,
      "learning_rate": 4.853210671587789e-05,
      "loss": 1.0824,
      "step": 3548
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12189384549856186,
      "learning_rate": 4.8478683900952274e-05,
      "loss": 1.0047,
      "step": 3549
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.13353689014911652,
      "learning_rate": 4.842528109631218e-05,
      "loss": 1.1272,
      "step": 3550
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12525786459445953,
      "learning_rate": 4.8371898322698585e-05,
      "loss": 1.1962,
      "step": 3551
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.15626177191734314,
      "learning_rate": 4.831853560084477e-05,
      "loss": 1.1331,
      "step": 3552
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.138467937707901,
      "learning_rate": 4.82651929514762e-05,
      "loss": 1.1062,
      "step": 3553
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.1319926232099533,
      "learning_rate": 4.821187039531056e-05,
      "loss": 1.0661,
      "step": 3554
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.13768108189105988,
      "learning_rate": 4.815856795305772e-05,
      "loss": 1.1436,
      "step": 3555
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.12815240025520325,
      "learning_rate": 4.810528564541969e-05,
      "loss": 1.0943,
      "step": 3556
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.10570608824491501,
      "learning_rate": 4.805202349309074e-05,
      "loss": 1.1832,
      "step": 3557
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.11902832239866257,
      "learning_rate": 4.79987815167573e-05,
      "loss": 1.0898,
      "step": 3558
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.13462671637535095,
      "learning_rate": 4.7945559737097834e-05,
      "loss": 1.0206,
      "step": 3559
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.14135798811912537,
      "learning_rate": 4.789235817478322e-05,
      "loss": 1.074,
      "step": 3560
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1605537384748459,
      "learning_rate": 4.7839176850476206e-05,
      "loss": 1.1366,
      "step": 3561
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.14430268108844757,
      "learning_rate": 4.778601578483187e-05,
      "loss": 1.0934,
      "step": 3562
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.13903813064098358,
      "learning_rate": 4.773287499849737e-05,
      "loss": 1.132,
      "step": 3563
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.17413783073425293,
      "learning_rate": 4.767975451211191e-05,
      "loss": 1.1123,
      "step": 3564
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.15286950767040253,
      "learning_rate": 4.7626654346306923e-05,
      "loss": 1.0118,
      "step": 3565
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.17561966180801392,
      "learning_rate": 4.757357452170588e-05,
      "loss": 1.1607,
      "step": 3566
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.14940422773361206,
      "learning_rate": 4.752051505892438e-05,
      "loss": 1.234,
      "step": 3567
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.10603979974985123,
      "learning_rate": 4.746747597857014e-05,
      "loss": 1.1723,
      "step": 3568
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.13778595626354218,
      "learning_rate": 4.741445730124288e-05,
      "loss": 1.1961,
      "step": 3569
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1565093994140625,
      "learning_rate": 4.736145904753445e-05,
      "loss": 1.1167,
      "step": 3570
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.18626897037029266,
      "learning_rate": 4.7308481238028765e-05,
      "loss": 1.0939,
      "step": 3571
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.16304536163806915,
      "learning_rate": 4.725552389330182e-05,
      "loss": 1.1411,
      "step": 3572
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1454816460609436,
      "learning_rate": 4.720258703392161e-05,
      "loss": 1.2623,
      "step": 3573
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1409173309803009,
      "learning_rate": 4.7149670680448256e-05,
      "loss": 1.1801,
      "step": 3574
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.12251760810613632,
      "learning_rate": 4.709677485343377e-05,
      "loss": 1.1391,
      "step": 3575
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.12791939079761505,
      "learning_rate": 4.704389957342237e-05,
      "loss": 1.0507,
      "step": 3576
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.12602195143699646,
      "learning_rate": 4.699104486095008e-05,
      "loss": 1.2276,
      "step": 3577
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.14594195783138275,
      "learning_rate": 4.6938210736545195e-05,
      "loss": 1.0612,
      "step": 3578
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.12904050946235657,
      "learning_rate": 4.688539722072786e-05,
      "loss": 1.1535,
      "step": 3579
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.14963281154632568,
      "learning_rate": 4.6832604334010164e-05,
      "loss": 1.1093,
      "step": 3580
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.11306113749742508,
      "learning_rate": 4.67798320968963e-05,
      "loss": 1.1872,
      "step": 3581
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.13039083778858185,
      "learning_rate": 4.672708052988239e-05,
      "loss": 1.1454,
      "step": 3582
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.15265467762947083,
      "learning_rate": 4.667434965345654e-05,
      "loss": 1.2022,
      "step": 3583
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1617046296596527,
      "learning_rate": 4.6621639488098854e-05,
      "loss": 1.058,
      "step": 3584
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.14728504419326782,
      "learning_rate": 4.656895005428127e-05,
      "loss": 1.1866,
      "step": 3585
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.20008762180805206,
      "learning_rate": 4.6516281372467806e-05,
      "loss": 1.0278,
      "step": 3586
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.14055553078651428,
      "learning_rate": 4.64636334631144e-05,
      "loss": 1.14,
      "step": 3587
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.11586879938840866,
      "learning_rate": 4.6411006346668776e-05,
      "loss": 1.0648,
      "step": 3588
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1384691298007965,
      "learning_rate": 4.635840004357086e-05,
      "loss": 1.0752,
      "step": 3589
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.11583945900201797,
      "learning_rate": 4.630581457425222e-05,
      "loss": 1.1399,
      "step": 3590
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.14432689547538757,
      "learning_rate": 4.625324995913648e-05,
      "loss": 1.1839,
      "step": 3591
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.15997782349586487,
      "learning_rate": 4.620070621863917e-05,
      "loss": 1.0675,
      "step": 3592
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1522606462240219,
      "learning_rate": 4.614818337316759e-05,
      "loss": 1.1181,
      "step": 3593
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.15310066938400269,
      "learning_rate": 4.6095681443121066e-05,
      "loss": 1.1545,
      "step": 3594
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.116330586373806,
      "learning_rate": 4.604320044889072e-05,
      "loss": 1.17,
      "step": 3595
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1228753924369812,
      "learning_rate": 4.599074041085958e-05,
      "loss": 1.1001,
      "step": 3596
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1525501161813736,
      "learning_rate": 4.593830134940256e-05,
      "loss": 1.1141,
      "step": 3597
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.15048523247241974,
      "learning_rate": 4.588588328488629e-05,
      "loss": 1.1499,
      "step": 3598
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.11863092333078384,
      "learning_rate": 4.583348623766941e-05,
      "loss": 1.1189,
      "step": 3599
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.13898517191410065,
      "learning_rate": 4.578111022810231e-05,
      "loss": 1.1691,
      "step": 3600
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.11979557573795319,
      "learning_rate": 4.5728755276527226e-05,
      "loss": 1.1782,
      "step": 3601
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.14383085072040558,
      "learning_rate": 4.567642140327823e-05,
      "loss": 1.1063,
      "step": 3602
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1162552759051323,
      "learning_rate": 4.562410862868123e-05,
      "loss": 1.1598,
      "step": 3603
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1442110240459442,
      "learning_rate": 4.5571816973053825e-05,
      "loss": 1.1549,
      "step": 3604
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.13941967487335205,
      "learning_rate": 4.551954645670557e-05,
      "loss": 1.1557,
      "step": 3605
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.10717065632343292,
      "learning_rate": 4.5467297099937625e-05,
      "loss": 1.1805,
      "step": 3606
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.11922302842140198,
      "learning_rate": 4.5415068923043133e-05,
      "loss": 1.0447,
      "step": 3607
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.12334927171468735,
      "learning_rate": 4.5362861946306936e-05,
      "loss": 1.1563,
      "step": 3608
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.12527117133140564,
      "learning_rate": 4.531067619000553e-05,
      "loss": 1.1809,
      "step": 3609
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.11095918715000153,
      "learning_rate": 4.525851167440731e-05,
      "loss": 1.0619,
      "step": 3610
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.14272469282150269,
      "learning_rate": 4.5206368419772406e-05,
      "loss": 1.0725,
      "step": 3611
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.16050179302692413,
      "learning_rate": 4.5154246446352544e-05,
      "loss": 1.0734,
      "step": 3612
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15646663308143616,
      "learning_rate": 4.5102145774391455e-05,
      "loss": 1.0972,
      "step": 3613
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.13804951310157776,
      "learning_rate": 4.505006642412433e-05,
      "loss": 1.1225,
      "step": 3614
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.0917016789317131,
      "learning_rate": 4.4998008415778205e-05,
      "loss": 1.0804,
      "step": 3615
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11972659081220627,
      "learning_rate": 4.494597176957186e-05,
      "loss": 1.1106,
      "step": 3616
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11043412983417511,
      "learning_rate": 4.4893956505715614e-05,
      "loss": 1.1876,
      "step": 3617
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.16147196292877197,
      "learning_rate": 4.484196264441176e-05,
      "loss": 1.1164,
      "step": 3618
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11110226809978485,
      "learning_rate": 4.478999020585399e-05,
      "loss": 1.1277,
      "step": 3619
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15632455050945282,
      "learning_rate": 4.4738039210227836e-05,
      "loss": 1.2173,
      "step": 3620
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11027960479259491,
      "learning_rate": 4.468610967771051e-05,
      "loss": 1.2382,
      "step": 3621
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.13919928669929504,
      "learning_rate": 4.463420162847076e-05,
      "loss": 1.098,
      "step": 3622
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1290637105703354,
      "learning_rate": 4.4582315082669124e-05,
      "loss": 1.1104,
      "step": 3623
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11770142614841461,
      "learning_rate": 4.453045006045773e-05,
      "loss": 1.1512,
      "step": 3624
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15682551264762878,
      "learning_rate": 4.4478606581980354e-05,
      "loss": 1.1739,
      "step": 3625
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11209271103143692,
      "learning_rate": 4.4426784667372444e-05,
      "loss": 1.1249,
      "step": 3626
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.14899905025959015,
      "learning_rate": 4.437498433676097e-05,
      "loss": 1.1707,
      "step": 3627
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15058311820030212,
      "learning_rate": 4.432320561026461e-05,
      "loss": 1.1357,
      "step": 3628
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.14767596125602722,
      "learning_rate": 4.4271448507993626e-05,
      "loss": 1.0843,
      "step": 3629
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.13988672196865082,
      "learning_rate": 4.4219713050049884e-05,
      "loss": 1.1355,
      "step": 3630
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15163911879062653,
      "learning_rate": 4.416799925652684e-05,
      "loss": 1.2112,
      "step": 3631
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11771151423454285,
      "learning_rate": 4.4116307147509564e-05,
      "loss": 1.1706,
      "step": 3632
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.12430921941995621,
      "learning_rate": 4.4064636743074605e-05,
      "loss": 1.1201,
      "step": 3633
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1312878131866455,
      "learning_rate": 4.4012988063290194e-05,
      "loss": 1.1936,
      "step": 3634
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11199118196964264,
      "learning_rate": 4.3961361128216084e-05,
      "loss": 1.0937,
      "step": 3635
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1507163792848587,
      "learning_rate": 4.390975595790358e-05,
      "loss": 1.1508,
      "step": 3636
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1261441707611084,
      "learning_rate": 4.385817257239556e-05,
      "loss": 1.0346,
      "step": 3637
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1472482830286026,
      "learning_rate": 4.380661099172636e-05,
      "loss": 1.1724,
      "step": 3638
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1082974374294281,
      "learning_rate": 4.375507123592194e-05,
      "loss": 1.1462,
      "step": 3639
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.13179227709770203,
      "learning_rate": 4.370355332499977e-05,
      "loss": 1.087,
      "step": 3640
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11642702668905258,
      "learning_rate": 4.365205727896872e-05,
      "loss": 1.2164,
      "step": 3641
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.13437162339687347,
      "learning_rate": 4.3600583117829395e-05,
      "loss": 1.1307,
      "step": 3642
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.14199919998645782,
      "learning_rate": 4.354913086157366e-05,
      "loss": 1.0963,
      "step": 3643
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15037277340888977,
      "learning_rate": 4.3497700530185015e-05,
      "loss": 1.1237,
      "step": 3644
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.14300420880317688,
      "learning_rate": 4.344629214363845e-05,
      "loss": 1.0856,
      "step": 3645
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.16836945712566376,
      "learning_rate": 4.33949057219003e-05,
      "loss": 1.1539,
      "step": 3646
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.16465547680854797,
      "learning_rate": 4.3343541284928515e-05,
      "loss": 1.1079,
      "step": 3647
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1456468403339386,
      "learning_rate": 4.329219885267244e-05,
      "loss": 1.1369,
      "step": 3648
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15441490709781647,
      "learning_rate": 4.324087844507289e-05,
      "loss": 1.1124,
      "step": 3649
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.10792689025402069,
      "learning_rate": 4.318958008206214e-05,
      "loss": 1.1059,
      "step": 3650
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.14171044528484344,
      "learning_rate": 4.3138303783563835e-05,
      "loss": 1.1171,
      "step": 3651
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11438456177711487,
      "learning_rate": 4.308704956949313e-05,
      "loss": 1.0761,
      "step": 3652
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1549489051103592,
      "learning_rate": 4.303581745975655e-05,
      "loss": 1.1357,
      "step": 3653
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15723560750484467,
      "learning_rate": 4.298460747425208e-05,
      "loss": 1.1078,
      "step": 3654
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.10721265524625778,
      "learning_rate": 4.293341963286912e-05,
      "loss": 1.1528,
      "step": 3655
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11134260147809982,
      "learning_rate": 4.288225395548835e-05,
      "loss": 1.2115,
      "step": 3656
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1976919025182724,
      "learning_rate": 4.283111046198198e-05,
      "loss": 1.0906,
      "step": 3657
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.13005392253398895,
      "learning_rate": 4.277998917221354e-05,
      "loss": 1.0883,
      "step": 3658
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.10874234884977341,
      "learning_rate": 4.272889010603798e-05,
      "loss": 1.0137,
      "step": 3659
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.11874790489673615,
      "learning_rate": 4.267781328330155e-05,
      "loss": 1.1264,
      "step": 3660
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.16038578748703003,
      "learning_rate": 4.262675872384196e-05,
      "loss": 1.0971,
      "step": 3661
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15967008471488953,
      "learning_rate": 4.257572644748813e-05,
      "loss": 1.2378,
      "step": 3662
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.15097610652446747,
      "learning_rate": 4.2524716474060454e-05,
      "loss": 1.1401,
      "step": 3663
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.12352873384952545,
      "learning_rate": 4.2473728823370604e-05,
      "loss": 1.1462,
      "step": 3664
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.13342510163784027,
      "learning_rate": 4.242276351522161e-05,
      "loss": 1.1548,
      "step": 3665
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.14956241846084595,
      "learning_rate": 4.237182056940784e-05,
      "loss": 1.1206,
      "step": 3666
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.14732477068901062,
      "learning_rate": 4.232090000571488e-05,
      "loss": 1.065,
      "step": 3667
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1610066294670105,
      "learning_rate": 4.227000184391972e-05,
      "loss": 1.1493,
      "step": 3668
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.16589903831481934,
      "learning_rate": 4.221912610379065e-05,
      "loss": 1.1299,
      "step": 3669
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12921126186847687,
      "learning_rate": 4.2168272805087114e-05,
      "loss": 1.1515,
      "step": 3670
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.11728020757436752,
      "learning_rate": 4.2117441967560115e-05,
      "loss": 1.1013,
      "step": 3671
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.14031724631786346,
      "learning_rate": 4.2066633610951636e-05,
      "loss": 1.1234,
      "step": 3672
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.15573370456695557,
      "learning_rate": 4.2015847754995086e-05,
      "loss": 1.1064,
      "step": 3673
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12938740849494934,
      "learning_rate": 4.196508441941516e-05,
      "loss": 1.0801,
      "step": 3674
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.13138943910598755,
      "learning_rate": 4.191434362392768e-05,
      "loss": 1.1371,
      "step": 3675
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1244533434510231,
      "learning_rate": 4.18636253882398e-05,
      "loss": 1.1635,
      "step": 3676
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1423613280057907,
      "learning_rate": 4.181292973204992e-05,
      "loss": 1.0664,
      "step": 3677
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.14746570587158203,
      "learning_rate": 4.176225667504765e-05,
      "loss": 1.1405,
      "step": 3678
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1373174786567688,
      "learning_rate": 4.171160623691383e-05,
      "loss": 1.1127,
      "step": 3679
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.11846545338630676,
      "learning_rate": 4.1660978437320474e-05,
      "loss": 1.1489,
      "step": 3680
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12068387866020203,
      "learning_rate": 4.161037329593085e-05,
      "loss": 1.0036,
      "step": 3681
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12753209471702576,
      "learning_rate": 4.155979083239942e-05,
      "loss": 1.1322,
      "step": 3682
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.14393998682498932,
      "learning_rate": 4.150923106637181e-05,
      "loss": 1.109,
      "step": 3683
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12383370101451874,
      "learning_rate": 4.145869401748491e-05,
      "loss": 1.0834,
      "step": 3684
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.15787416696548462,
      "learning_rate": 4.1408179705366636e-05,
      "loss": 1.0638,
      "step": 3685
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.14242303371429443,
      "learning_rate": 4.135768814963622e-05,
      "loss": 1.069,
      "step": 3686
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1319640725851059,
      "learning_rate": 4.130721936990399e-05,
      "loss": 1.1156,
      "step": 3687
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1497504562139511,
      "learning_rate": 4.125677338577144e-05,
      "loss": 1.1602,
      "step": 3688
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.11289259046316147,
      "learning_rate": 4.1206350216831214e-05,
      "loss": 1.0406,
      "step": 3689
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.16737200319766998,
      "learning_rate": 4.115594988266711e-05,
      "loss": 1.1115,
      "step": 3690
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12442764639854431,
      "learning_rate": 4.110557240285398e-05,
      "loss": 1.1145,
      "step": 3691
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1292956918478012,
      "learning_rate": 4.105521779695789e-05,
      "loss": 1.1574,
      "step": 3692
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.20106863975524902,
      "learning_rate": 4.100488608453599e-05,
      "loss": 1.0592,
      "step": 3693
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1694418042898178,
      "learning_rate": 4.095457728513652e-05,
      "loss": 1.0834,
      "step": 3694
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.13632215559482574,
      "learning_rate": 4.09042914182989e-05,
      "loss": 1.1631,
      "step": 3695
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.14845028519630432,
      "learning_rate": 4.08540285035535e-05,
      "loss": 1.0658,
      "step": 3696
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.10958141088485718,
      "learning_rate": 4.0803788560421885e-05,
      "loss": 1.1172,
      "step": 3697
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12590201199054718,
      "learning_rate": 4.075357160841671e-05,
      "loss": 1.0896,
      "step": 3698
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.11853957176208496,
      "learning_rate": 4.0703377667041554e-05,
      "loss": 1.1752,
      "step": 3699
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.14175114035606384,
      "learning_rate": 4.065320675579132e-05,
      "loss": 1.1311,
      "step": 3700
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.11596961319446564,
      "learning_rate": 4.060305889415168e-05,
      "loss": 1.1293,
      "step": 3701
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12344858795404434,
      "learning_rate": 4.055293410159954e-05,
      "loss": 1.0995,
      "step": 3702
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.15910597145557404,
      "learning_rate": 4.050283239760282e-05,
      "loss": 1.1152,
      "step": 3703
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.16344323754310608,
      "learning_rate": 4.045275380162038e-05,
      "loss": 1.1886,
      "step": 3704
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.13475312292575836,
      "learning_rate": 4.04026983331022e-05,
      "loss": 1.2035,
      "step": 3705
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.16002792119979858,
      "learning_rate": 4.035266601148924e-05,
      "loss": 1.2474,
      "step": 3706
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1194402351975441,
      "learning_rate": 4.030265685621349e-05,
      "loss": 1.0932,
      "step": 3707
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.13667911291122437,
      "learning_rate": 4.025267088669796e-05,
      "loss": 1.1727,
      "step": 3708
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.13008500635623932,
      "learning_rate": 4.020270812235656e-05,
      "loss": 1.1178,
      "step": 3709
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.13012123107910156,
      "learning_rate": 4.015276858259427e-05,
      "loss": 1.1751,
      "step": 3710
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12874777615070343,
      "learning_rate": 4.0102852286807044e-05,
      "loss": 1.1675,
      "step": 3711
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.10595318675041199,
      "learning_rate": 4.005295925438181e-05,
      "loss": 1.1323,
      "step": 3712
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.13486729562282562,
      "learning_rate": 4.000308950469646e-05,
      "loss": 1.1479,
      "step": 3713
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.12501755356788635,
      "learning_rate": 3.995324305711976e-05,
      "loss": 1.1279,
      "step": 3714
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1366562396287918,
      "learning_rate": 3.990341993101154e-05,
      "loss": 1.1306,
      "step": 3715
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.14388270676136017,
      "learning_rate": 3.985362014572256e-05,
      "loss": 1.1173,
      "step": 3716
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1434464454650879,
      "learning_rate": 3.980384372059438e-05,
      "loss": 1.1528,
      "step": 3717
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.13892190158367157,
      "learning_rate": 3.97540906749597e-05,
      "loss": 1.1668,
      "step": 3718
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.11648226529359818,
      "learning_rate": 3.9704361028142025e-05,
      "loss": 1.1712,
      "step": 3719
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1362370252609253,
      "learning_rate": 3.9654654799455695e-05,
      "loss": 1.0723,
      "step": 3720
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1742171198129654,
      "learning_rate": 3.960497200820609e-05,
      "loss": 1.1219,
      "step": 3721
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.13068078458309174,
      "learning_rate": 3.955531267368941e-05,
      "loss": 1.1048,
      "step": 3722
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.12406259030103683,
      "learning_rate": 3.9505676815192794e-05,
      "loss": 1.1709,
      "step": 3723
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.13223250210285187,
      "learning_rate": 3.945606445199427e-05,
      "loss": 1.1056,
      "step": 3724
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1704278290271759,
      "learning_rate": 3.9406475603362617e-05,
      "loss": 1.1541,
      "step": 3725
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1266642063856125,
      "learning_rate": 3.9356910288557626e-05,
      "loss": 1.1277,
      "step": 3726
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.16191378235816956,
      "learning_rate": 3.9307368526829924e-05,
      "loss": 1.0356,
      "step": 3727
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1354941576719284,
      "learning_rate": 3.925785033742085e-05,
      "loss": 1.1877,
      "step": 3728
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.142691507935524,
      "learning_rate": 3.920835573956285e-05,
      "loss": 1.0678,
      "step": 3729
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.08701815456151962,
      "learning_rate": 3.915888475247894e-05,
      "loss": 1.146,
      "step": 3730
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.14137502014636993,
      "learning_rate": 3.910943739538313e-05,
      "loss": 1.127,
      "step": 3731
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1313449889421463,
      "learning_rate": 3.906001368748022e-05,
      "loss": 1.159,
      "step": 3732
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1480160653591156,
      "learning_rate": 3.901061364796574e-05,
      "loss": 1.0625,
      "step": 3733
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.13668131828308105,
      "learning_rate": 3.896123729602615e-05,
      "loss": 1.1307,
      "step": 3734
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.14422360062599182,
      "learning_rate": 3.891188465083865e-05,
      "loss": 1.1183,
      "step": 3735
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.11753267049789429,
      "learning_rate": 3.8862555731571216e-05,
      "loss": 1.1578,
      "step": 3736
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.15188685059547424,
      "learning_rate": 3.88132505573827e-05,
      "loss": 1.1564,
      "step": 3737
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.14942950010299683,
      "learning_rate": 3.876396914742257e-05,
      "loss": 1.1128,
      "step": 3738
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.14061367511749268,
      "learning_rate": 3.87147115208312e-05,
      "loss": 1.183,
      "step": 3739
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.23502019047737122,
      "learning_rate": 3.8665477696739685e-05,
      "loss": 1.1328,
      "step": 3740
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.13920362293720245,
      "learning_rate": 3.861626769426988e-05,
      "loss": 1.1372,
      "step": 3741
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1363631784915924,
      "learning_rate": 3.856708153253438e-05,
      "loss": 1.0875,
      "step": 3742
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.12708324193954468,
      "learning_rate": 3.8517919230636546e-05,
      "loss": 1.1561,
      "step": 3743
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.11499596387147903,
      "learning_rate": 3.846878080767039e-05,
      "loss": 1.1834,
      "step": 3744
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.09748668968677521,
      "learning_rate": 3.841966628272079e-05,
      "loss": 1.1482,
      "step": 3745
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.13930146396160126,
      "learning_rate": 3.837057567486314e-05,
      "loss": 1.0709,
      "step": 3746
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.10965407639741898,
      "learning_rate": 3.832150900316377e-05,
      "loss": 1.1404,
      "step": 3747
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.13274413347244263,
      "learning_rate": 3.827246628667962e-05,
      "loss": 1.1098,
      "step": 3748
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.16671644151210785,
      "learning_rate": 3.8223447544458256e-05,
      "loss": 1.1791,
      "step": 3749
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1860247105360031,
      "learning_rate": 3.817445279553801e-05,
      "loss": 1.0928,
      "step": 3750
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.15024980902671814,
      "learning_rate": 3.8125482058947906e-05,
      "loss": 1.1952,
      "step": 3751
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.14126896858215332,
      "learning_rate": 3.807653535370752e-05,
      "loss": 1.1338,
      "step": 3752
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1512867659330368,
      "learning_rate": 3.802761269882734e-05,
      "loss": 1.0958,
      "step": 3753
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.1289878785610199,
      "learning_rate": 3.797871411330824e-05,
      "loss": 1.1653,
      "step": 3754
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.10951082408428192,
      "learning_rate": 3.7929839616141914e-05,
      "loss": 1.0552,
      "step": 3755
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.14489586651325226,
      "learning_rate": 3.788098922631067e-05,
      "loss": 1.0487,
      "step": 3756
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.15329064428806305,
      "learning_rate": 3.7832162962787355e-05,
      "loss": 1.2089,
      "step": 3757
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.15295414626598358,
      "learning_rate": 3.7783360844535656e-05,
      "loss": 1.1411,
      "step": 3758
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.15714576840400696,
      "learning_rate": 3.773458289050963e-05,
      "loss": 1.15,
      "step": 3759
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.11370337754487991,
      "learning_rate": 3.7685829119654134e-05,
      "loss": 1.1067,
      "step": 3760
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.18690426647663116,
      "learning_rate": 3.763709955090461e-05,
      "loss": 1.0714,
      "step": 3761
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.16026705503463745,
      "learning_rate": 3.758839420318696e-05,
      "loss": 1.1078,
      "step": 3762
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.14606358110904694,
      "learning_rate": 3.753971309541784e-05,
      "loss": 1.125,
      "step": 3763
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.14475412666797638,
      "learning_rate": 3.7491056246504433e-05,
      "loss": 1.1334,
      "step": 3764
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.11942199617624283,
      "learning_rate": 3.7442423675344474e-05,
      "loss": 1.2282,
      "step": 3765
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.12915277481079102,
      "learning_rate": 3.739381540082635e-05,
      "loss": 1.1548,
      "step": 3766
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.11404864490032196,
      "learning_rate": 3.734523144182887e-05,
      "loss": 1.0527,
      "step": 3767
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.12474334239959717,
      "learning_rate": 3.729667181722154e-05,
      "loss": 1.1851,
      "step": 3768
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.13004712760448456,
      "learning_rate": 3.7248136545864344e-05,
      "loss": 1.0967,
      "step": 3769
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.17080117762088776,
      "learning_rate": 3.7199625646607825e-05,
      "loss": 1.1905,
      "step": 3770
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.13860441744327545,
      "learning_rate": 3.7151139138293054e-05,
      "loss": 1.1673,
      "step": 3771
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.11150813847780228,
      "learning_rate": 3.7102677039751666e-05,
      "loss": 1.1829,
      "step": 3772
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.15537621080875397,
      "learning_rate": 3.705423936980572e-05,
      "loss": 1.0822,
      "step": 3773
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.13811075687408447,
      "learning_rate": 3.700582614726791e-05,
      "loss": 1.1737,
      "step": 3774
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.17301522195339203,
      "learning_rate": 3.6957437390941275e-05,
      "loss": 1.2475,
      "step": 3775
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.15453986823558807,
      "learning_rate": 3.690907311961955e-05,
      "loss": 1.0748,
      "step": 3776
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.11253956705331802,
      "learning_rate": 3.6860733352086864e-05,
      "loss": 1.1627,
      "step": 3777
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1322152465581894,
      "learning_rate": 3.681241810711776e-05,
      "loss": 1.1204,
      "step": 3778
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.11763017624616623,
      "learning_rate": 3.6764127403477344e-05,
      "loss": 1.0826,
      "step": 3779
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.159868523478508,
      "learning_rate": 3.671586125992123e-05,
      "loss": 1.0605,
      "step": 3780
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.16282601654529572,
      "learning_rate": 3.6667619695195285e-05,
      "loss": 1.0507,
      "step": 3781
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1868576854467392,
      "learning_rate": 3.661940272803616e-05,
      "loss": 1.0132,
      "step": 3782
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1401725560426712,
      "learning_rate": 3.657121037717064e-05,
      "loss": 1.0646,
      "step": 3783
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.14599357545375824,
      "learning_rate": 3.652304266131612e-05,
      "loss": 1.0849,
      "step": 3784
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.10955840349197388,
      "learning_rate": 3.6474899599180423e-05,
      "loss": 1.2161,
      "step": 3785
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1368500292301178,
      "learning_rate": 3.642678120946168e-05,
      "loss": 1.1028,
      "step": 3786
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.11520792543888092,
      "learning_rate": 3.6378687510848575e-05,
      "loss": 1.0736,
      "step": 3787
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.13158218562602997,
      "learning_rate": 3.633061852202012e-05,
      "loss": 1.0916,
      "step": 3788
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.11742973327636719,
      "learning_rate": 3.628257426164577e-05,
      "loss": 1.1289,
      "step": 3789
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.14674706757068634,
      "learning_rate": 3.62345547483854e-05,
      "loss": 1.12,
      "step": 3790
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.10436704754829407,
      "learning_rate": 3.6186560000889155e-05,
      "loss": 1.1591,
      "step": 3791
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.14214017987251282,
      "learning_rate": 3.613859003779769e-05,
      "loss": 1.0726,
      "step": 3792
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.12311791628599167,
      "learning_rate": 3.609064487774199e-05,
      "loss": 1.1548,
      "step": 3793
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.15182814002037048,
      "learning_rate": 3.6042724539343376e-05,
      "loss": 1.1297,
      "step": 3794
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1353846937417984,
      "learning_rate": 3.599482904121361e-05,
      "loss": 1.1705,
      "step": 3795
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1293911635875702,
      "learning_rate": 3.594695840195468e-05,
      "loss": 1.193,
      "step": 3796
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.15578100085258484,
      "learning_rate": 3.589911264015902e-05,
      "loss": 1.0915,
      "step": 3797
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.12704969942569733,
      "learning_rate": 3.585129177440938e-05,
      "loss": 1.1644,
      "step": 3798
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.12605519592761993,
      "learning_rate": 3.580349582327882e-05,
      "loss": 1.0848,
      "step": 3799
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.11155972629785538,
      "learning_rate": 3.575572480533076e-05,
      "loss": 1.1119,
      "step": 3800
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1263992041349411,
      "learning_rate": 3.570797873911892e-05,
      "loss": 1.1149,
      "step": 3801
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.15762373805046082,
      "learning_rate": 3.5660257643187276e-05,
      "loss": 1.1208,
      "step": 3802
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.14969412982463837,
      "learning_rate": 3.561256153607021e-05,
      "loss": 1.127,
      "step": 3803
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.15859781205654144,
      "learning_rate": 3.556489043629224e-05,
      "loss": 1.2176,
      "step": 3804
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.14441432058811188,
      "learning_rate": 3.5517244362368365e-05,
      "loss": 1.0738,
      "step": 3805
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.15853850543498993,
      "learning_rate": 3.546962333280379e-05,
      "loss": 1.139,
      "step": 3806
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1436820775270462,
      "learning_rate": 3.542202736609389e-05,
      "loss": 1.1466,
      "step": 3807
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.13957515358924866,
      "learning_rate": 3.5374456480724426e-05,
      "loss": 1.1747,
      "step": 3808
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.14356543123722076,
      "learning_rate": 3.532691069517142e-05,
      "loss": 1.104,
      "step": 3809
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1318843960762024,
      "learning_rate": 3.527939002790101e-05,
      "loss": 1.0977,
      "step": 3810
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.11282504349946976,
      "learning_rate": 3.52318944973698e-05,
      "loss": 1.0935,
      "step": 3811
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.13914576172828674,
      "learning_rate": 3.518442412202441e-05,
      "loss": 1.158,
      "step": 3812
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.1962774395942688,
      "learning_rate": 3.5136978920301825e-05,
      "loss": 1.0491,
      "step": 3813
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.17340055108070374,
      "learning_rate": 3.508955891062924e-05,
      "loss": 1.2067,
      "step": 3814
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.19418096542358398,
      "learning_rate": 3.504216411142398e-05,
      "loss": 1.1444,
      "step": 3815
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.12949404120445251,
      "learning_rate": 3.499479454109367e-05,
      "loss": 1.0397,
      "step": 3816
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.14442946016788483,
      "learning_rate": 3.4947450218036106e-05,
      "loss": 1.0525,
      "step": 3817
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.12129819393157959,
      "learning_rate": 3.490013116063928e-05,
      "loss": 1.1553,
      "step": 3818
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.13155585527420044,
      "learning_rate": 3.485283738728139e-05,
      "loss": 1.0173,
      "step": 3819
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.14936433732509613,
      "learning_rate": 3.480556891633074e-05,
      "loss": 1.0818,
      "step": 3820
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.143156036734581,
      "learning_rate": 3.475832576614589e-05,
      "loss": 1.0917,
      "step": 3821
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.129116028547287,
      "learning_rate": 3.4711107955075536e-05,
      "loss": 1.0992,
      "step": 3822
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1175515428185463,
      "learning_rate": 3.466391550145852e-05,
      "loss": 1.0591,
      "step": 3823
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.13004373013973236,
      "learning_rate": 3.461674842362389e-05,
      "loss": 1.0959,
      "step": 3824
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1419183313846588,
      "learning_rate": 3.456960673989074e-05,
      "loss": 1.1263,
      "step": 3825
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12138883769512177,
      "learning_rate": 3.452249046856836e-05,
      "loss": 1.1121,
      "step": 3826
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.13059432804584503,
      "learning_rate": 3.447539962795619e-05,
      "loss": 1.1181,
      "step": 3827
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.13028717041015625,
      "learning_rate": 3.442833423634377e-05,
      "loss": 1.0897,
      "step": 3828
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1240011677145958,
      "learning_rate": 3.4381294312010745e-05,
      "loss": 1.1626,
      "step": 3829
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.15530972182750702,
      "learning_rate": 3.433427987322693e-05,
      "loss": 1.0953,
      "step": 3830
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.10594236105680466,
      "learning_rate": 3.4287290938252106e-05,
      "loss": 1.093,
      "step": 3831
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12387064099311829,
      "learning_rate": 3.424032752533627e-05,
      "loss": 1.0304,
      "step": 3832
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.17023402452468872,
      "learning_rate": 3.4193389652719476e-05,
      "loss": 1.0779,
      "step": 3833
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12288490682840347,
      "learning_rate": 3.414647733863185e-05,
      "loss": 1.1399,
      "step": 3834
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.11915712803602219,
      "learning_rate": 3.4099590601293626e-05,
      "loss": 1.1292,
      "step": 3835
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2121439278125763,
      "learning_rate": 3.4052729458915023e-05,
      "loss": 1.1218,
      "step": 3836
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.10853013396263123,
      "learning_rate": 3.4005893929696377e-05,
      "loss": 1.1404,
      "step": 3837
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.10491643846035004,
      "learning_rate": 3.395908403182811e-05,
      "loss": 1.0554,
      "step": 3838
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12723346054553986,
      "learning_rate": 3.391229978349056e-05,
      "loss": 1.1499,
      "step": 3839
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1360282152891159,
      "learning_rate": 3.386554120285431e-05,
      "loss": 1.1388,
      "step": 3840
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.11717604100704193,
      "learning_rate": 3.381880830807975e-05,
      "loss": 1.1561,
      "step": 3841
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12245111912488937,
      "learning_rate": 3.3772101117317435e-05,
      "loss": 1.0782,
      "step": 3842
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.16450609266757965,
      "learning_rate": 3.372541964870795e-05,
      "loss": 1.1467,
      "step": 3843
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.10864174365997314,
      "learning_rate": 3.367876392038174e-05,
      "loss": 1.2503,
      "step": 3844
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1109033152461052,
      "learning_rate": 3.363213395045941e-05,
      "loss": 1.1081,
      "step": 3845
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12653428316116333,
      "learning_rate": 3.3585529757051504e-05,
      "loss": 1.1133,
      "step": 3846
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1354968249797821,
      "learning_rate": 3.353895135825854e-05,
      "loss": 1.0799,
      "step": 3847
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.155600443482399,
      "learning_rate": 3.349239877217108e-05,
      "loss": 1.0909,
      "step": 3848
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1348017156124115,
      "learning_rate": 3.3445872016869516e-05,
      "loss": 1.1566,
      "step": 3849
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.11759112030267715,
      "learning_rate": 3.339937111042437e-05,
      "loss": 1.0898,
      "step": 3850
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.11502469331026077,
      "learning_rate": 3.3352896070896054e-05,
      "loss": 1.1408,
      "step": 3851
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1438870131969452,
      "learning_rate": 3.330644691633492e-05,
      "loss": 1.1219,
      "step": 3852
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1854158192873001,
      "learning_rate": 3.3260023664781326e-05,
      "loss": 1.1494,
      "step": 3853
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.14530648291110992,
      "learning_rate": 3.321362633426547e-05,
      "loss": 1.1215,
      "step": 3854
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12687812745571136,
      "learning_rate": 3.316725494280757e-05,
      "loss": 1.1583,
      "step": 3855
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1408337652683258,
      "learning_rate": 3.312090950841775e-05,
      "loss": 1.0686,
      "step": 3856
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1450773924589157,
      "learning_rate": 3.307459004909599e-05,
      "loss": 1.1901,
      "step": 3857
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12007907778024673,
      "learning_rate": 3.3028296582832284e-05,
      "loss": 1.1327,
      "step": 3858
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12662379443645477,
      "learning_rate": 3.298202912760652e-05,
      "loss": 1.2177,
      "step": 3859
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.11969132721424103,
      "learning_rate": 3.2935787701388344e-05,
      "loss": 1.1374,
      "step": 3860
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.16810928285121918,
      "learning_rate": 3.288957232213745e-05,
      "loss": 1.1165,
      "step": 3861
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.10800683498382568,
      "learning_rate": 3.284338300780336e-05,
      "loss": 1.0639,
      "step": 3862
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.12342584133148193,
      "learning_rate": 3.279721977632546e-05,
      "loss": 1.1045,
      "step": 3863
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.16668589413166046,
      "learning_rate": 3.275108264563306e-05,
      "loss": 1.1472,
      "step": 3864
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.13355940580368042,
      "learning_rate": 3.270497163364521e-05,
      "loss": 1.11,
      "step": 3865
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2425745576620102,
      "learning_rate": 3.2658886758270943e-05,
      "loss": 1.0873,
      "step": 3866
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.13344165682792664,
      "learning_rate": 3.261282803740911e-05,
      "loss": 1.1497,
      "step": 3867
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.13461121916770935,
      "learning_rate": 3.256679548894831e-05,
      "loss": 1.0794,
      "step": 3868
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.162210151553154,
      "learning_rate": 3.252078913076718e-05,
      "loss": 1.0905,
      "step": 3869
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.13245396316051483,
      "learning_rate": 3.247480898073395e-05,
      "loss": 1.1317,
      "step": 3870
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.11574536561965942,
      "learning_rate": 3.242885505670681e-05,
      "loss": 1.0815,
      "step": 3871
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.15647003054618835,
      "learning_rate": 3.238292737653379e-05,
      "loss": 1.0877,
      "step": 3872
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.10130614787340164,
      "learning_rate": 3.233702595805258e-05,
      "loss": 1.1683,
      "step": 3873
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13232728838920593,
      "learning_rate": 3.229115081909082e-05,
      "loss": 1.1516,
      "step": 3874
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13362392783164978,
      "learning_rate": 3.224530197746587e-05,
      "loss": 1.0825,
      "step": 3875
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.12727244198322296,
      "learning_rate": 3.219947945098489e-05,
      "loss": 1.1262,
      "step": 3876
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.12174341827630997,
      "learning_rate": 3.215368325744485e-05,
      "loss": 1.1298,
      "step": 3877
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.14494001865386963,
      "learning_rate": 3.2107913414632426e-05,
      "loss": 1.1446,
      "step": 3878
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1010410338640213,
      "learning_rate": 3.206216994032411e-05,
      "loss": 1.1666,
      "step": 3879
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1264081597328186,
      "learning_rate": 3.2016452852286127e-05,
      "loss": 1.1527,
      "step": 3880
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.15814544260501862,
      "learning_rate": 3.197076216827449e-05,
      "loss": 1.0712,
      "step": 3881
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.14320531487464905,
      "learning_rate": 3.192509790603496e-05,
      "loss": 1.2048,
      "step": 3882
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1368008553981781,
      "learning_rate": 3.187946008330295e-05,
      "loss": 1.0774,
      "step": 3883
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1151009351015091,
      "learning_rate": 3.183384871780367e-05,
      "loss": 1.202,
      "step": 3884
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13876873254776,
      "learning_rate": 3.178826382725212e-05,
      "loss": 1.0356,
      "step": 3885
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1977718323469162,
      "learning_rate": 3.1742705429352826e-05,
      "loss": 1.119,
      "step": 3886
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.12316666543483734,
      "learning_rate": 3.169717354180025e-05,
      "loss": 1.0093,
      "step": 3887
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.12251655012369156,
      "learning_rate": 3.165166818227845e-05,
      "loss": 1.1261,
      "step": 3888
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13897691667079926,
      "learning_rate": 3.160618936846111e-05,
      "loss": 1.0908,
      "step": 3889
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.10578544437885284,
      "learning_rate": 3.156073711801172e-05,
      "loss": 1.1167,
      "step": 3890
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.14446516335010529,
      "learning_rate": 3.151531144858344e-05,
      "loss": 1.163,
      "step": 3891
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.14610818028450012,
      "learning_rate": 3.1469912377818986e-05,
      "loss": 1.0896,
      "step": 3892
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.17028945684432983,
      "learning_rate": 3.142453992335096e-05,
      "loss": 1.023,
      "step": 3893
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1406768262386322,
      "learning_rate": 3.137919410280139e-05,
      "loss": 1.1524,
      "step": 3894
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13344039022922516,
      "learning_rate": 3.1333874933782115e-05,
      "loss": 1.1055,
      "step": 3895
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1323913186788559,
      "learning_rate": 3.128858243389461e-05,
      "loss": 1.1124,
      "step": 3896
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.12217836827039719,
      "learning_rate": 3.124331662072987e-05,
      "loss": 1.0589,
      "step": 3897
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13071520626544952,
      "learning_rate": 3.119807751186872e-05,
      "loss": 1.1048,
      "step": 3898
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1268025040626526,
      "learning_rate": 3.115286512488144e-05,
      "loss": 1.0487,
      "step": 3899
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1719488650560379,
      "learning_rate": 3.110767947732801e-05,
      "loss": 1.2098,
      "step": 3900
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13558532297611237,
      "learning_rate": 3.106252058675806e-05,
      "loss": 1.3028,
      "step": 3901
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.14335763454437256,
      "learning_rate": 3.1017388470710716e-05,
      "loss": 1.113,
      "step": 3902
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.17092950642108917,
      "learning_rate": 3.097228314671481e-05,
      "loss": 1.1263,
      "step": 3903
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.12353041768074036,
      "learning_rate": 3.092720463228872e-05,
      "loss": 1.1153,
      "step": 3904
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.12990519404411316,
      "learning_rate": 3.0882152944940424e-05,
      "loss": 1.1297,
      "step": 3905
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.14928853511810303,
      "learning_rate": 3.0837128102167515e-05,
      "loss": 1.155,
      "step": 3906
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.14205172657966614,
      "learning_rate": 3.079213012145705e-05,
      "loss": 1.1735,
      "step": 3907
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.17139695584774017,
      "learning_rate": 3.0747159020285765e-05,
      "loss": 1.0907,
      "step": 3908
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13897623121738434,
      "learning_rate": 3.070221481611992e-05,
      "loss": 1.1849,
      "step": 3909
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13042064011096954,
      "learning_rate": 3.065729752641532e-05,
      "loss": 1.1965,
      "step": 3910
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.12902554869651794,
      "learning_rate": 3.061240716861735e-05,
      "loss": 1.0848,
      "step": 3911
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13119624555110931,
      "learning_rate": 3.056754376016087e-05,
      "loss": 1.2131,
      "step": 3912
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.11043515056371689,
      "learning_rate": 3.05227073184703e-05,
      "loss": 1.0496,
      "step": 3913
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.14938345551490784,
      "learning_rate": 3.047789786095967e-05,
      "loss": 1.1029,
      "step": 3914
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.11416371166706085,
      "learning_rate": 3.043311540503233e-05,
      "loss": 1.1666,
      "step": 3915
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.17600978910923004,
      "learning_rate": 3.0388359968081392e-05,
      "loss": 1.097,
      "step": 3916
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1487978994846344,
      "learning_rate": 3.0343631567489327e-05,
      "loss": 1.1304,
      "step": 3917
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.12983767688274384,
      "learning_rate": 3.0298930220628086e-05,
      "loss": 1.1503,
      "step": 3918
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13590025901794434,
      "learning_rate": 3.0254255944859187e-05,
      "loss": 1.207,
      "step": 3919
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1266484260559082,
      "learning_rate": 3.0209608757533625e-05,
      "loss": 1.2186,
      "step": 3920
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.11686329543590546,
      "learning_rate": 3.0164988675991768e-05,
      "loss": 1.175,
      "step": 3921
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13888943195343018,
      "learning_rate": 3.0120395717563653e-05,
      "loss": 1.104,
      "step": 3922
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1358700394630432,
      "learning_rate": 3.0075829899568597e-05,
      "loss": 1.1911,
      "step": 3923
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.13461947441101074,
      "learning_rate": 3.0031291239315475e-05,
      "loss": 1.2545,
      "step": 3924
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.10939455032348633,
      "learning_rate": 2.9986779754102613e-05,
      "loss": 1.205,
      "step": 3925
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.12572844326496124,
      "learning_rate": 2.9942295461217696e-05,
      "loss": 1.1586,
      "step": 3926
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.11223046481609344,
      "learning_rate": 2.9897838377937947e-05,
      "loss": 1.0837,
      "step": 3927
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1411411613225937,
      "learning_rate": 2.985340852152999e-05,
      "loss": 1.0654,
      "step": 3928
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.17672695219516754,
      "learning_rate": 2.9809005909249864e-05,
      "loss": 1.1092,
      "step": 3929
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.16573433578014374,
      "learning_rate": 2.9764630558343067e-05,
      "loss": 1.1263,
      "step": 3930
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13823693990707397,
      "learning_rate": 2.9720282486044403e-05,
      "loss": 1.2342,
      "step": 3931
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.14820054173469543,
      "learning_rate": 2.9675961709578194e-05,
      "loss": 1.1715,
      "step": 3932
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.11411868035793304,
      "learning_rate": 2.9631668246158107e-05,
      "loss": 1.0589,
      "step": 3933
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.12137552350759506,
      "learning_rate": 2.958740211298722e-05,
      "loss": 1.1367,
      "step": 3934
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.16169556975364685,
      "learning_rate": 2.9543163327258017e-05,
      "loss": 1.1584,
      "step": 3935
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.15127314627170563,
      "learning_rate": 2.9498951906152272e-05,
      "loss": 1.1488,
      "step": 3936
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.11934265494346619,
      "learning_rate": 2.9454767866841227e-05,
      "loss": 1.0407,
      "step": 3937
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.14474929869174957,
      "learning_rate": 2.9410611226485452e-05,
      "loss": 1.1556,
      "step": 3938
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1615191549062729,
      "learning_rate": 2.9366482002234874e-05,
      "loss": 1.0398,
      "step": 3939
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1519196480512619,
      "learning_rate": 2.932238021122877e-05,
      "loss": 1.2253,
      "step": 3940
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.16111759841442108,
      "learning_rate": 2.9278305870595812e-05,
      "loss": 1.09,
      "step": 3941
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.15691284835338593,
      "learning_rate": 2.92342589974539e-05,
      "loss": 1.0712,
      "step": 3942
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13595065474510193,
      "learning_rate": 2.9190239608910387e-05,
      "loss": 1.123,
      "step": 3943
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.14146529138088226,
      "learning_rate": 2.9146247722061802e-05,
      "loss": 1.0627,
      "step": 3944
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13298529386520386,
      "learning_rate": 2.9102283353994186e-05,
      "loss": 1.108,
      "step": 3945
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13177724182605743,
      "learning_rate": 2.9058346521782797e-05,
      "loss": 1.1357,
      "step": 3946
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.10501198470592499,
      "learning_rate": 2.9014437242492132e-05,
      "loss": 1.0849,
      "step": 3947
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.11434558033943176,
      "learning_rate": 2.897055553317607e-05,
      "loss": 1.0756,
      "step": 3948
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1227881982922554,
      "learning_rate": 2.8926701410877798e-05,
      "loss": 1.0619,
      "step": 3949
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.12861548364162445,
      "learning_rate": 2.8882874892629652e-05,
      "loss": 1.104,
      "step": 3950
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.14561989903450012,
      "learning_rate": 2.883907599545348e-05,
      "loss": 1.0969,
      "step": 3951
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1471307873725891,
      "learning_rate": 2.8795304736360184e-05,
      "loss": 1.0057,
      "step": 3952
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13276535272598267,
      "learning_rate": 2.8751561132350024e-05,
      "loss": 1.0531,
      "step": 3953
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.12477774918079376,
      "learning_rate": 2.8707845200412565e-05,
      "loss": 1.1804,
      "step": 3954
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13587120175361633,
      "learning_rate": 2.866415695752649e-05,
      "loss": 1.1687,
      "step": 3955
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1194266602396965,
      "learning_rate": 2.862049642065986e-05,
      "loss": 1.2401,
      "step": 3956
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.11354994773864746,
      "learning_rate": 2.857686360676991e-05,
      "loss": 1.1196,
      "step": 3957
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1290186047554016,
      "learning_rate": 2.853325853280312e-05,
      "loss": 1.1379,
      "step": 3958
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.12418042868375778,
      "learning_rate": 2.848968121569524e-05,
      "loss": 1.1301,
      "step": 3959
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1412416398525238,
      "learning_rate": 2.8446131672371135e-05,
      "loss": 1.0033,
      "step": 3960
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.10736475139856339,
      "learning_rate": 2.8402609919744972e-05,
      "loss": 1.1534,
      "step": 3961
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2072005718946457,
      "learning_rate": 2.8359115974720096e-05,
      "loss": 1.0619,
      "step": 3962
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1255311518907547,
      "learning_rate": 2.8315649854189062e-05,
      "loss": 1.2123,
      "step": 3963
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.11920582503080368,
      "learning_rate": 2.8272211575033636e-05,
      "loss": 1.1526,
      "step": 3964
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13985861837863922,
      "learning_rate": 2.8228801154124685e-05,
      "loss": 1.1344,
      "step": 3965
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.11289297044277191,
      "learning_rate": 2.8185418608322344e-05,
      "loss": 1.1301,
      "step": 3966
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.15498410165309906,
      "learning_rate": 2.8142063954475927e-05,
      "loss": 1.1611,
      "step": 3967
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.17452767491340637,
      "learning_rate": 2.8098737209423797e-05,
      "loss": 1.1301,
      "step": 3968
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13668952882289886,
      "learning_rate": 2.8055438389993637e-05,
      "loss": 1.1722,
      "step": 3969
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.11376068741083145,
      "learning_rate": 2.801216751300223e-05,
      "loss": 1.1156,
      "step": 3970
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13114263117313385,
      "learning_rate": 2.7968924595255407e-05,
      "loss": 1.1023,
      "step": 3971
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.11737065762281418,
      "learning_rate": 2.792570965354829e-05,
      "loss": 1.1778,
      "step": 3972
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.1445125788450241,
      "learning_rate": 2.788252270466497e-05,
      "loss": 1.1142,
      "step": 3973
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13730183243751526,
      "learning_rate": 2.7839363765378858e-05,
      "loss": 1.1155,
      "step": 3974
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.15840108692646027,
      "learning_rate": 2.7796232852452377e-05,
      "loss": 1.0844,
      "step": 3975
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.13700847327709198,
      "learning_rate": 2.7753129982637026e-05,
      "loss": 1.1962,
      "step": 3976
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12021582573652267,
      "learning_rate": 2.7710055172673488e-05,
      "loss": 1.0797,
      "step": 3977
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12482719123363495,
      "learning_rate": 2.7667008439291554e-05,
      "loss": 1.1124,
      "step": 3978
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.14462654292583466,
      "learning_rate": 2.762398979920998e-05,
      "loss": 1.0903,
      "step": 3979
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.15283234417438507,
      "learning_rate": 2.7580999269136855e-05,
      "loss": 1.0769,
      "step": 3980
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12338368594646454,
      "learning_rate": 2.753803686576909e-05,
      "loss": 1.1347,
      "step": 3981
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1466784030199051,
      "learning_rate": 2.7495102605792823e-05,
      "loss": 1.1459,
      "step": 3982
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12568631768226624,
      "learning_rate": 2.7452196505883265e-05,
      "loss": 1.0753,
      "step": 3983
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12065508216619492,
      "learning_rate": 2.740931858270459e-05,
      "loss": 1.0398,
      "step": 3984
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.11788754165172577,
      "learning_rate": 2.736646885291011e-05,
      "loss": 1.1824,
      "step": 3985
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.14297594130039215,
      "learning_rate": 2.7323647333142177e-05,
      "loss": 1.168,
      "step": 3986
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.13903099298477173,
      "learning_rate": 2.7280854040032165e-05,
      "loss": 1.1391,
      "step": 3987
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12098846584558487,
      "learning_rate": 2.723808899020054e-05,
      "loss": 1.1594,
      "step": 3988
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1717412918806076,
      "learning_rate": 2.7195352200256674e-05,
      "loss": 1.1158,
      "step": 3989
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.20366068184375763,
      "learning_rate": 2.7152643686799094e-05,
      "loss": 1.0332,
      "step": 3990
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.0977993905544281,
      "learning_rate": 2.710996346641528e-05,
      "loss": 1.0549,
      "step": 3991
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.10345537960529327,
      "learning_rate": 2.7067311555681753e-05,
      "loss": 1.135,
      "step": 3992
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.14000727236270905,
      "learning_rate": 2.702468797116403e-05,
      "loss": 1.0391,
      "step": 3993
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1311149299144745,
      "learning_rate": 2.6982092729416587e-05,
      "loss": 1.1113,
      "step": 3994
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12985564768314362,
      "learning_rate": 2.693952584698294e-05,
      "loss": 1.1108,
      "step": 3995
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.14954274892807007,
      "learning_rate": 2.6896987340395607e-05,
      "loss": 1.1008,
      "step": 3996
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.14459124207496643,
      "learning_rate": 2.6854477226175966e-05,
      "loss": 1.153,
      "step": 3997
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1297638714313507,
      "learning_rate": 2.6811995520834542e-05,
      "loss": 1.1594,
      "step": 3998
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.10590071231126785,
      "learning_rate": 2.676954224087075e-05,
      "loss": 1.0602,
      "step": 3999
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12407078593969345,
      "learning_rate": 2.6727117402772884e-05,
      "loss": 1.0753,
      "step": 4000
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1160527914762497,
      "learning_rate": 2.668472102301829e-05,
      "loss": 1.0975,
      "step": 4001
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1208387166261673,
      "learning_rate": 2.664235311807327e-05,
      "loss": 1.1854,
      "step": 4002
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.15107278525829315,
      "learning_rate": 2.6600013704392946e-05,
      "loss": 1.084,
      "step": 4003
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.15040720999240875,
      "learning_rate": 2.6557702798421568e-05,
      "loss": 1.1232,
      "step": 4004
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1561790555715561,
      "learning_rate": 2.6515420416592106e-05,
      "loss": 1.2032,
      "step": 4005
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.16185034811496735,
      "learning_rate": 2.6473166575326604e-05,
      "loss": 1.1445,
      "step": 4006
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.14251476526260376,
      "learning_rate": 2.643094129103598e-05,
      "loss": 1.1273,
      "step": 4007
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.16366177797317505,
      "learning_rate": 2.6388744580119974e-05,
      "loss": 1.1542,
      "step": 4008
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.14828433096408844,
      "learning_rate": 2.6346576458967398e-05,
      "loss": 1.1461,
      "step": 4009
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.11572246253490448,
      "learning_rate": 2.630443694395579e-05,
      "loss": 1.1319,
      "step": 4010
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.10143768787384033,
      "learning_rate": 2.626232605145168e-05,
      "loss": 1.1109,
      "step": 4011
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.16112221777439117,
      "learning_rate": 2.6220243797810485e-05,
      "loss": 1.1204,
      "step": 4012
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.14177271723747253,
      "learning_rate": 2.6178190199376396e-05,
      "loss": 1.1626,
      "step": 4013
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.10736539959907532,
      "learning_rate": 2.6136165272482594e-05,
      "loss": 1.1397,
      "step": 4014
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.11387249827384949,
      "learning_rate": 2.6094169033451066e-05,
      "loss": 1.1648,
      "step": 4015
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.13010385632514954,
      "learning_rate": 2.6052201498592667e-05,
      "loss": 1.099,
      "step": 4016
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.10164612531661987,
      "learning_rate": 2.6010262684207133e-05,
      "loss": 1.0016,
      "step": 4017
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.19796352088451385,
      "learning_rate": 2.596835260658297e-05,
      "loss": 1.1401,
      "step": 4018
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.15724991261959076,
      "learning_rate": 2.5926471281997577e-05,
      "loss": 1.1607,
      "step": 4019
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12507592141628265,
      "learning_rate": 2.588461872671719e-05,
      "loss": 1.1945,
      "step": 4020
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1463761329650879,
      "learning_rate": 2.5842794956996865e-05,
      "loss": 1.1351,
      "step": 4021
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.13438907265663147,
      "learning_rate": 2.5800999989080487e-05,
      "loss": 1.0807,
      "step": 4022
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2180289775133133,
      "learning_rate": 2.575923383920069e-05,
      "loss": 1.2699,
      "step": 4023
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12323862314224243,
      "learning_rate": 2.5717496523578998e-05,
      "loss": 1.1751,
      "step": 4024
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12502256035804749,
      "learning_rate": 2.567578805842572e-05,
      "loss": 1.1198,
      "step": 4025
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1719626933336258,
      "learning_rate": 2.5634108459939877e-05,
      "loss": 1.0743,
      "step": 4026
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12404169887304306,
      "learning_rate": 2.5592457744309404e-05,
      "loss": 1.0594,
      "step": 4027
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.12854339182376862,
      "learning_rate": 2.555083592771098e-05,
      "loss": 1.0261,
      "step": 4028
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.15852144360542297,
      "learning_rate": 2.5509243026309982e-05,
      "loss": 1.1615,
      "step": 4029
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.12160142511129379,
      "learning_rate": 2.546767905626063e-05,
      "loss": 1.0369,
      "step": 4030
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.13874393701553345,
      "learning_rate": 2.5426144033705935e-05,
      "loss": 1.1353,
      "step": 4031
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1236959844827652,
      "learning_rate": 2.5384637974777515e-05,
      "loss": 1.0892,
      "step": 4032
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1257353127002716,
      "learning_rate": 2.5343160895595974e-05,
      "loss": 1.0983,
      "step": 4033
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.12593042850494385,
      "learning_rate": 2.530171281227044e-05,
      "loss": 1.1286,
      "step": 4034
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.12230115383863449,
      "learning_rate": 2.52602937408989e-05,
      "loss": 1.15,
      "step": 4035
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.12576256692409515,
      "learning_rate": 2.5218903697568076e-05,
      "loss": 1.0941,
      "step": 4036
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.13578182458877563,
      "learning_rate": 2.5177542698353317e-05,
      "loss": 1.1231,
      "step": 4037
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1109374463558197,
      "learning_rate": 2.5136210759318812e-05,
      "loss": 1.1792,
      "step": 4038
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.16348572075366974,
      "learning_rate": 2.5094907896517383e-05,
      "loss": 1.1799,
      "step": 4039
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.13377337157726288,
      "learning_rate": 2.5053634125990587e-05,
      "loss": 1.1234,
      "step": 4040
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.14923225343227386,
      "learning_rate": 2.5012389463768737e-05,
      "loss": 1.2112,
      "step": 4041
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.194574773311615,
      "learning_rate": 2.4971173925870693e-05,
      "loss": 1.1606,
      "step": 4042
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.15456314384937286,
      "learning_rate": 2.4929987528304144e-05,
      "loss": 1.1268,
      "step": 4043
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1566878855228424,
      "learning_rate": 2.4888830287065412e-05,
      "loss": 1.1273,
      "step": 4044
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.10586079955101013,
      "learning_rate": 2.484770221813949e-05,
      "loss": 1.0944,
      "step": 4045
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.18549250066280365,
      "learning_rate": 2.4806603337500067e-05,
      "loss": 1.1036,
      "step": 4046
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.13712993264198303,
      "learning_rate": 2.476553366110944e-05,
      "loss": 1.0035,
      "step": 4047
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.13379287719726562,
      "learning_rate": 2.4724493204918596e-05,
      "loss": 1.0775,
      "step": 4048
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.14124801754951477,
      "learning_rate": 2.4683481984867207e-05,
      "loss": 1.1387,
      "step": 4049
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.11756888031959534,
      "learning_rate": 2.4642500016883528e-05,
      "loss": 1.0675,
      "step": 4050
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.14143671095371246,
      "learning_rate": 2.4601547316884543e-05,
      "loss": 1.1852,
      "step": 4051
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.11750835180282593,
      "learning_rate": 2.4560623900775727e-05,
      "loss": 1.0598,
      "step": 4052
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.11824418604373932,
      "learning_rate": 2.4519729784451295e-05,
      "loss": 1.1341,
      "step": 4053
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.12187172472476959,
      "learning_rate": 2.4478864983794093e-05,
      "loss": 1.1504,
      "step": 4054
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.12612730264663696,
      "learning_rate": 2.443802951467544e-05,
      "loss": 1.1334,
      "step": 4055
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1345996856689453,
      "learning_rate": 2.4397223392955447e-05,
      "loss": 1.104,
      "step": 4056
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1659008264541626,
      "learning_rate": 2.4356446634482754e-05,
      "loss": 1.0318,
      "step": 4057
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.10411630570888519,
      "learning_rate": 2.4315699255094515e-05,
      "loss": 1.0812,
      "step": 4058
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1253357231616974,
      "learning_rate": 2.4274981270616583e-05,
      "loss": 1.1248,
      "step": 4059
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1412547379732132,
      "learning_rate": 2.4234292696863358e-05,
      "loss": 1.1179,
      "step": 4060
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.13829505443572998,
      "learning_rate": 2.419363354963776e-05,
      "loss": 1.1823,
      "step": 4061
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.12358442693948746,
      "learning_rate": 2.4153003844731425e-05,
      "loss": 1.0976,
      "step": 4062
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.138805091381073,
      "learning_rate": 2.4112403597924384e-05,
      "loss": 1.1436,
      "step": 4063
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.13551582396030426,
      "learning_rate": 2.407183282498534e-05,
      "loss": 1.1004,
      "step": 4064
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1602981686592102,
      "learning_rate": 2.403129154167153e-05,
      "loss": 1.1337,
      "step": 4065
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.13318713009357452,
      "learning_rate": 2.3990779763728666e-05,
      "loss": 1.1473,
      "step": 4066
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.15856844186782837,
      "learning_rate": 2.3950297506891084e-05,
      "loss": 1.185,
      "step": 4067
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.12924031913280487,
      "learning_rate": 2.390984478688164e-05,
      "loss": 1.1102,
      "step": 4068
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.17300765216350555,
      "learning_rate": 2.386942161941169e-05,
      "loss": 1.148,
      "step": 4069
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.14742250740528107,
      "learning_rate": 2.3829028020181155e-05,
      "loss": 1.058,
      "step": 4070
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.15460440516471863,
      "learning_rate": 2.37886640048784e-05,
      "loss": 1.0934,
      "step": 4071
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.117367222905159,
      "learning_rate": 2.374832958918035e-05,
      "loss": 1.1816,
      "step": 4072
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1421480029821396,
      "learning_rate": 2.370802478875245e-05,
      "loss": 1.1503,
      "step": 4073
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.126296728849411,
      "learning_rate": 2.3667749619248614e-05,
      "loss": 1.0685,
      "step": 4074
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.15020476281642914,
      "learning_rate": 2.362750409631127e-05,
      "loss": 1.1905,
      "step": 4075
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1313004493713379,
      "learning_rate": 2.3587288235571258e-05,
      "loss": 1.0987,
      "step": 4076
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.11349713057279587,
      "learning_rate": 2.3547102052648006e-05,
      "loss": 1.053,
      "step": 4077
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.11562524735927582,
      "learning_rate": 2.350694556314934e-05,
      "loss": 1.0805,
      "step": 4078
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.1493973582983017,
      "learning_rate": 2.3466818782671596e-05,
      "loss": 1.072,
      "step": 4079
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.13031548261642456,
      "learning_rate": 2.3426721726799573e-05,
      "loss": 1.2005,
      "step": 4080
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1194462850689888,
      "learning_rate": 2.3386654411106447e-05,
      "loss": 1.1989,
      "step": 4081
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.15038995444774628,
      "learning_rate": 2.3346616851153933e-05,
      "loss": 1.1799,
      "step": 4082
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.11002760380506516,
      "learning_rate": 2.330660906249218e-05,
      "loss": 1.1791,
      "step": 4083
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.11910390853881836,
      "learning_rate": 2.3266631060659682e-05,
      "loss": 1.1837,
      "step": 4084
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.13415955007076263,
      "learning_rate": 2.3226682861183503e-05,
      "loss": 1.1141,
      "step": 4085
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.11645203083753586,
      "learning_rate": 2.318676447957907e-05,
      "loss": 1.1349,
      "step": 4086
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.14018407464027405,
      "learning_rate": 2.3146875931350166e-05,
      "loss": 1.1214,
      "step": 4087
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1492149382829666,
      "learning_rate": 2.310701723198908e-05,
      "loss": 1.1215,
      "step": 4088
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.14210136234760284,
      "learning_rate": 2.306718839697648e-05,
      "loss": 1.1109,
      "step": 4089
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.11340191215276718,
      "learning_rate": 2.3027389441781366e-05,
      "loss": 1.1355,
      "step": 4090
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.15764586627483368,
      "learning_rate": 2.2987620381861285e-05,
      "loss": 1.1162,
      "step": 4091
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.14362627267837524,
      "learning_rate": 2.2947881232662006e-05,
      "loss": 1.0901,
      "step": 4092
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.11265746504068375,
      "learning_rate": 2.290817200961779e-05,
      "loss": 1.0474,
      "step": 4093
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.11078481376171112,
      "learning_rate": 2.2868492728151258e-05,
      "loss": 1.1288,
      "step": 4094
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1323656439781189,
      "learning_rate": 2.282884340367334e-05,
      "loss": 1.0831,
      "step": 4095
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.13651634752750397,
      "learning_rate": 2.2789224051583403e-05,
      "loss": 1.1583,
      "step": 4096
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.12849794328212738,
      "learning_rate": 2.274963468726914e-05,
      "loss": 1.121,
      "step": 4097
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.13827191293239594,
      "learning_rate": 2.2710075326106617e-05,
      "loss": 1.103,
      "step": 4098
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.11589039117097855,
      "learning_rate": 2.2670545983460243e-05,
      "loss": 1.1061,
      "step": 4099
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1336400955915451,
      "learning_rate": 2.263104667468272e-05,
      "loss": 1.1157,
      "step": 4100
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1685895025730133,
      "learning_rate": 2.259157741511515e-05,
      "loss": 1.1113,
      "step": 4101
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.12508653104305267,
      "learning_rate": 2.2552138220086927e-05,
      "loss": 1.1143,
      "step": 4102
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.15272031724452972,
      "learning_rate": 2.2512729104915786e-05,
      "loss": 1.1328,
      "step": 4103
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1901167333126068,
      "learning_rate": 2.2473350084907806e-05,
      "loss": 1.0212,
      "step": 4104
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.18112562596797943,
      "learning_rate": 2.243400117535729e-05,
      "loss": 1.099,
      "step": 4105
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.16905029118061066,
      "learning_rate": 2.2394682391546928e-05,
      "loss": 1.0661,
      "step": 4106
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.13761648535728455,
      "learning_rate": 2.23553937487477e-05,
      "loss": 1.1167,
      "step": 4107
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.11956635117530823,
      "learning_rate": 2.2316135262218785e-05,
      "loss": 1.1409,
      "step": 4108
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.188825786113739,
      "learning_rate": 2.2276906947207844e-05,
      "loss": 1.0615,
      "step": 4109
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.14720021188259125,
      "learning_rate": 2.2237708818950607e-05,
      "loss": 1.0959,
      "step": 4110
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1529446691274643,
      "learning_rate": 2.2198540892671215e-05,
      "loss": 1.1049,
      "step": 4111
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1325504332780838,
      "learning_rate": 2.215940318358206e-05,
      "loss": 1.2118,
      "step": 4112
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.12919385731220245,
      "learning_rate": 2.2120295706883698e-05,
      "loss": 1.1562,
      "step": 4113
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.14269833266735077,
      "learning_rate": 2.20812184777651e-05,
      "loss": 1.1309,
      "step": 4114
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.156743124127388,
      "learning_rate": 2.204217151140342e-05,
      "loss": 1.1429,
      "step": 4115
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.15743473172187805,
      "learning_rate": 2.2003154822963978e-05,
      "loss": 1.1447,
      "step": 4116
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.12401141226291656,
      "learning_rate": 2.196416842760046e-05,
      "loss": 1.1707,
      "step": 4117
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.15517757833003998,
      "learning_rate": 2.1925212340454736e-05,
      "loss": 1.1006,
      "step": 4118
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1212153434753418,
      "learning_rate": 2.1886286576656835e-05,
      "loss": 1.1252,
      "step": 4119
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.12489887326955795,
      "learning_rate": 2.184739115132517e-05,
      "loss": 1.0802,
      "step": 4120
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1577915996313095,
      "learning_rate": 2.1808526079566217e-05,
      "loss": 1.0943,
      "step": 4121
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.16801896691322327,
      "learning_rate": 2.176969137647472e-05,
      "loss": 1.112,
      "step": 4122
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.13527697324752808,
      "learning_rate": 2.1730887057133677e-05,
      "loss": 1.2314,
      "step": 4123
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.16433140635490417,
      "learning_rate": 2.1692113136614177e-05,
      "loss": 1.0613,
      "step": 4124
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.13611677289009094,
      "learning_rate": 2.1653369629975595e-05,
      "loss": 1.1418,
      "step": 4125
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.15248610079288483,
      "learning_rate": 2.1614656552265456e-05,
      "loss": 1.1324,
      "step": 4126
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1514562964439392,
      "learning_rate": 2.1575973918519486e-05,
      "loss": 1.0916,
      "step": 4127
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1196976900100708,
      "learning_rate": 2.1537321743761584e-05,
      "loss": 1.2686,
      "step": 4128
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1330672651529312,
      "learning_rate": 2.1498700043003773e-05,
      "loss": 1.1129,
      "step": 4129
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.19495157897472382,
      "learning_rate": 2.1460108831246296e-05,
      "loss": 1.0608,
      "step": 4130
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1481180340051651,
      "learning_rate": 2.1421548123477532e-05,
      "loss": 1.1423,
      "step": 4131
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.13542412221431732,
      "learning_rate": 2.138301793467401e-05,
      "loss": 1.1742,
      "step": 4132
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12649884819984436,
      "learning_rate": 2.1344518279800452e-05,
      "loss": 1.1113,
      "step": 4133
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12576083838939667,
      "learning_rate": 2.1306049173809615e-05,
      "loss": 1.1314,
      "step": 4134
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12678563594818115,
      "learning_rate": 2.1267610631642498e-05,
      "loss": 1.0986,
      "step": 4135
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.15332023799419403,
      "learning_rate": 2.1229202668228197e-05,
      "loss": 1.1242,
      "step": 4136
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14134058356285095,
      "learning_rate": 2.1190825298483852e-05,
      "loss": 1.0782,
      "step": 4137
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12616339325904846,
      "learning_rate": 2.115247853731488e-05,
      "loss": 1.01,
      "step": 4138
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.13653524219989777,
      "learning_rate": 2.11141623996147e-05,
      "loss": 1.1085,
      "step": 4139
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.13893207907676697,
      "learning_rate": 2.107587690026481e-05,
      "loss": 1.1316,
      "step": 4140
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1168050616979599,
      "learning_rate": 2.103762205413493e-05,
      "loss": 1.1196,
      "step": 4141
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1396084576845169,
      "learning_rate": 2.0999397876082728e-05,
      "loss": 1.0299,
      "step": 4142
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.11339163035154343,
      "learning_rate": 2.0961204380954036e-05,
      "loss": 1.0891,
      "step": 4143
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.15880215167999268,
      "learning_rate": 2.092304158358286e-05,
      "loss": 1.1004,
      "step": 4144
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.18345744907855988,
      "learning_rate": 2.0884909498791104e-05,
      "loss": 1.1225,
      "step": 4145
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1565881222486496,
      "learning_rate": 2.084680814138885e-05,
      "loss": 1.1311,
      "step": 4146
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14313039183616638,
      "learning_rate": 2.080873752617426e-05,
      "loss": 1.094,
      "step": 4147
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12780176103115082,
      "learning_rate": 2.0770697667933437e-05,
      "loss": 1.0446,
      "step": 4148
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.11674383282661438,
      "learning_rate": 2.073268858144074e-05,
      "loss": 1.1156,
      "step": 4149
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.20917055010795593,
      "learning_rate": 2.0694710281458373e-05,
      "loss": 1.1303,
      "step": 4150
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.16316157579421997,
      "learning_rate": 2.0656762782736693e-05,
      "loss": 1.129,
      "step": 4151
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.13247686624526978,
      "learning_rate": 2.0618846100014112e-05,
      "loss": 1.1343,
      "step": 4152
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14008523523807526,
      "learning_rate": 2.058096024801697e-05,
      "loss": 1.1171,
      "step": 4153
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.132583349943161,
      "learning_rate": 2.0543105241459715e-05,
      "loss": 1.1631,
      "step": 4154
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14506296813488007,
      "learning_rate": 2.0505281095044804e-05,
      "loss": 1.1053,
      "step": 4155
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12583573162555695,
      "learning_rate": 2.0467487823462695e-05,
      "loss": 1.1324,
      "step": 4156
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14016945660114288,
      "learning_rate": 2.0429725441391888e-05,
      "loss": 1.2312,
      "step": 4157
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14820541441440582,
      "learning_rate": 2.0391993963498813e-05,
      "loss": 1.095,
      "step": 4158
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.10199877619743347,
      "learning_rate": 2.0354293404437965e-05,
      "loss": 1.145,
      "step": 4159
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1204652264714241,
      "learning_rate": 2.0316623778851783e-05,
      "loss": 1.1192,
      "step": 4160
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.15558144450187683,
      "learning_rate": 2.027898510137075e-05,
      "loss": 1.0792,
      "step": 4161
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.11873693764209747,
      "learning_rate": 2.024137738661329e-05,
      "loss": 1.1611,
      "step": 4162
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.16947080194950104,
      "learning_rate": 2.0203800649185788e-05,
      "loss": 1.1219,
      "step": 4163
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.13884668052196503,
      "learning_rate": 2.0166254903682603e-05,
      "loss": 1.1187,
      "step": 4164
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14306122064590454,
      "learning_rate": 2.0128740164686134e-05,
      "loss": 1.182,
      "step": 4165
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.11335566639900208,
      "learning_rate": 2.009125644676656e-05,
      "loss": 1.0841,
      "step": 4166
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.11237074434757233,
      "learning_rate": 2.0053803764482227e-05,
      "loss": 1.0947,
      "step": 4167
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12730039656162262,
      "learning_rate": 2.0016382132379318e-05,
      "loss": 1.1773,
      "step": 4168
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14757324755191803,
      "learning_rate": 1.9978991564991913e-05,
      "loss": 1.1291,
      "step": 4169
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12548784911632538,
      "learning_rate": 1.994163207684212e-05,
      "loss": 1.1438,
      "step": 4170
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14273041486740112,
      "learning_rate": 1.9904303682439897e-05,
      "loss": 1.0123,
      "step": 4171
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.13463163375854492,
      "learning_rate": 1.986700639628316e-05,
      "loss": 1.1961,
      "step": 4172
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.17171728610992432,
      "learning_rate": 1.9829740232857808e-05,
      "loss": 1.167,
      "step": 4173
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.17164357006549835,
      "learning_rate": 1.9792505206637524e-05,
      "loss": 1.2086,
      "step": 4174
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12000155448913574,
      "learning_rate": 1.9755301332083997e-05,
      "loss": 1.1032,
      "step": 4175
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1507796347141266,
      "learning_rate": 1.971812862364679e-05,
      "loss": 1.1746,
      "step": 4176
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14173348248004913,
      "learning_rate": 1.9680987095763313e-05,
      "loss": 1.0943,
      "step": 4177
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.15296855568885803,
      "learning_rate": 1.9643876762858937e-05,
      "loss": 1.2146,
      "step": 4178
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.14997299015522003,
      "learning_rate": 1.9606797639346874e-05,
      "loss": 1.1911,
      "step": 4179
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12181005626916885,
      "learning_rate": 1.9569749739628242e-05,
      "loss": 1.0955,
      "step": 4180
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.12193337827920914,
      "learning_rate": 1.9532733078092037e-05,
      "loss": 1.1688,
      "step": 4181
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.16635611653327942,
      "learning_rate": 1.949574766911506e-05,
      "loss": 1.0733,
      "step": 4182
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.15408365428447723,
      "learning_rate": 1.9458793527062035e-05,
      "loss": 1.1903,
      "step": 4183
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.11044152081012726,
      "learning_rate": 1.9421870666285524e-05,
      "loss": 1.2037,
      "step": 4184
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.11759164929389954,
      "learning_rate": 1.9384979101125943e-05,
      "loss": 1.1446,
      "step": 4185
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14796122908592224,
      "learning_rate": 1.934811884591159e-05,
      "loss": 1.1769,
      "step": 4186
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.12801595032215118,
      "learning_rate": 1.9311289914958497e-05,
      "loss": 1.1071,
      "step": 4187
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.11103743314743042,
      "learning_rate": 1.9274492322570615e-05,
      "loss": 1.2046,
      "step": 4188
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1149730458855629,
      "learning_rate": 1.923772608303972e-05,
      "loss": 1.1204,
      "step": 4189
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.16499236226081848,
      "learning_rate": 1.920099121064539e-05,
      "loss": 1.0834,
      "step": 4190
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.11085904389619827,
      "learning_rate": 1.9164287719655062e-05,
      "loss": 1.1155,
      "step": 4191
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14832767844200134,
      "learning_rate": 1.912761562432388e-05,
      "loss": 1.0853,
      "step": 4192
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.12020311504602432,
      "learning_rate": 1.90909749388949e-05,
      "loss": 1.1617,
      "step": 4193
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.15839266777038574,
      "learning_rate": 1.905436567759896e-05,
      "loss": 1.0554,
      "step": 4194
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14588218927383423,
      "learning_rate": 1.901778785465461e-05,
      "loss": 1.108,
      "step": 4195
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14280062913894653,
      "learning_rate": 1.898124148426832e-05,
      "loss": 1.0995,
      "step": 4196
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1599593311548233,
      "learning_rate": 1.8944726580634288e-05,
      "loss": 1.1242,
      "step": 4197
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.19527868926525116,
      "learning_rate": 1.8908243157934423e-05,
      "loss": 1.0825,
      "step": 4198
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1714356392621994,
      "learning_rate": 1.8871791230338497e-05,
      "loss": 1.1293,
      "step": 4199
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14166708290576935,
      "learning_rate": 1.883537081200404e-05,
      "loss": 1.1027,
      "step": 4200
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14773783087730408,
      "learning_rate": 1.8798981917076252e-05,
      "loss": 1.1579,
      "step": 4201
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.17897096276283264,
      "learning_rate": 1.8762624559688256e-05,
      "loss": 1.0852,
      "step": 4202
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.12394177913665771,
      "learning_rate": 1.872629875396076e-05,
      "loss": 1.0837,
      "step": 4203
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.12424508482217789,
      "learning_rate": 1.8690004514002313e-05,
      "loss": 1.1216,
      "step": 4204
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.12521454691886902,
      "learning_rate": 1.86537418539092e-05,
      "loss": 1.1376,
      "step": 4205
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.25732484459877014,
      "learning_rate": 1.861751078776538e-05,
      "loss": 1.1065,
      "step": 4206
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14277464151382446,
      "learning_rate": 1.858131132964259e-05,
      "loss": 1.1214,
      "step": 4207
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.08835700154304504,
      "learning_rate": 1.8545143493600292e-05,
      "loss": 1.0776,
      "step": 4208
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.11119592934846878,
      "learning_rate": 1.8509007293685667e-05,
      "loss": 1.1554,
      "step": 4209
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1513538360595703,
      "learning_rate": 1.8472902743933607e-05,
      "loss": 1.0998,
      "step": 4210
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1518583744764328,
      "learning_rate": 1.8436829858366657e-05,
      "loss": 1.1035,
      "step": 4211
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14449606835842133,
      "learning_rate": 1.840078865099514e-05,
      "loss": 1.0764,
      "step": 4212
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.173917755484581,
      "learning_rate": 1.8364779135817044e-05,
      "loss": 1.1375,
      "step": 4213
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.13641192018985748,
      "learning_rate": 1.8328801326818046e-05,
      "loss": 0.9618,
      "step": 4214
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.15400844812393188,
      "learning_rate": 1.829285523797155e-05,
      "loss": 1.1873,
      "step": 4215
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.17616192996501923,
      "learning_rate": 1.8256940883238538e-05,
      "loss": 1.1166,
      "step": 4216
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.12748992443084717,
      "learning_rate": 1.822105827656776e-05,
      "loss": 1.0621,
      "step": 4217
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1496627777814865,
      "learning_rate": 1.818520743189561e-05,
      "loss": 1.146,
      "step": 4218
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.1156248152256012,
      "learning_rate": 1.814938836314615e-05,
      "loss": 1.1267,
      "step": 4219
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.0972864106297493,
      "learning_rate": 1.8113601084231092e-05,
      "loss": 1.1111,
      "step": 4220
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.11318039894104004,
      "learning_rate": 1.8077845609049782e-05,
      "loss": 1.1241,
      "step": 4221
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.09625890105962753,
      "learning_rate": 1.8042121951489256e-05,
      "loss": 1.0921,
      "step": 4222
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.17348234355449677,
      "learning_rate": 1.800643012542418e-05,
      "loss": 1.174,
      "step": 4223
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14423561096191406,
      "learning_rate": 1.7970770144716774e-05,
      "loss": 1.1814,
      "step": 4224
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.12376614660024643,
      "learning_rate": 1.7935142023217057e-05,
      "loss": 1.1366,
      "step": 4225
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.16235829889774323,
      "learning_rate": 1.789954577476257e-05,
      "loss": 1.1223,
      "step": 4226
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14968328177928925,
      "learning_rate": 1.786398141317843e-05,
      "loss": 1.0919,
      "step": 4227
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.12391389161348343,
      "learning_rate": 1.7828448952277453e-05,
      "loss": 1.1089,
      "step": 4228
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.11447332799434662,
      "learning_rate": 1.7792948405860077e-05,
      "loss": 1.1212,
      "step": 4229
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.17977698147296906,
      "learning_rate": 1.7757479787714217e-05,
      "loss": 1.0579,
      "step": 4230
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.13181674480438232,
      "learning_rate": 1.7722043111615573e-05,
      "loss": 1.1136,
      "step": 4231
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.15149378776550293,
      "learning_rate": 1.7686638391327272e-05,
      "loss": 1.1408,
      "step": 4232
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.17341946065425873,
      "learning_rate": 1.765126564060011e-05,
      "loss": 1.0802,
      "step": 4233
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14022624492645264,
      "learning_rate": 1.7615924873172507e-05,
      "loss": 1.0513,
      "step": 4234
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.14435337483882904,
      "learning_rate": 1.7580616102770354e-05,
      "loss": 1.0715,
      "step": 4235
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.127250537276268,
      "learning_rate": 1.754533934310717e-05,
      "loss": 1.1239,
      "step": 4236
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.14101257920265198,
      "learning_rate": 1.7510094607884074e-05,
      "loss": 1.0965,
      "step": 4237
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.11362546682357788,
      "learning_rate": 1.74748819107897e-05,
      "loss": 1.1686,
      "step": 4238
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.16404418647289276,
      "learning_rate": 1.7439701265500273e-05,
      "loss": 1.0071,
      "step": 4239
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.15362492203712463,
      "learning_rate": 1.7404552685679508e-05,
      "loss": 1.1199,
      "step": 4240
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.11749681830406189,
      "learning_rate": 1.7369436184978736e-05,
      "loss": 1.1637,
      "step": 4241
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.14111635088920593,
      "learning_rate": 1.7334351777036805e-05,
      "loss": 1.1398,
      "step": 4242
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.12927372753620148,
      "learning_rate": 1.729929947548008e-05,
      "loss": 1.0896,
      "step": 4243
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1376645863056183,
      "learning_rate": 1.7264279293922502e-05,
      "loss": 1.0955,
      "step": 4244
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1259019523859024,
      "learning_rate": 1.7229291245965462e-05,
      "loss": 1.1619,
      "step": 4245
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.15069982409477234,
      "learning_rate": 1.7194335345197932e-05,
      "loss": 1.1257,
      "step": 4246
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.11827121675014496,
      "learning_rate": 1.715941160519641e-05,
      "loss": 1.1641,
      "step": 4247
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.11624453216791153,
      "learning_rate": 1.7124520039524803e-05,
      "loss": 1.1408,
      "step": 4248
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.17098848521709442,
      "learning_rate": 1.7089660661734685e-05,
      "loss": 1.1237,
      "step": 4249
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.13595689833164215,
      "learning_rate": 1.705483348536496e-05,
      "loss": 1.1537,
      "step": 4250
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1537141352891922,
      "learning_rate": 1.702003852394214e-05,
      "loss": 1.1455,
      "step": 4251
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1361575871706009,
      "learning_rate": 1.6985275790980203e-05,
      "loss": 1.1206,
      "step": 4252
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.11894833296537399,
      "learning_rate": 1.6950545299980526e-05,
      "loss": 1.173,
      "step": 4253
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.13417020440101624,
      "learning_rate": 1.691584706443209e-05,
      "loss": 1.0824,
      "step": 4254
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.13845863938331604,
      "learning_rate": 1.6881181097811304e-05,
      "loss": 1.0831,
      "step": 4255
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.13238687813282013,
      "learning_rate": 1.684654741358198e-05,
      "loss": 1.1115,
      "step": 4256
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.13069622218608856,
      "learning_rate": 1.681194602519546e-05,
      "loss": 1.1441,
      "step": 4257
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.15754270553588867,
      "learning_rate": 1.677737694609055e-05,
      "loss": 1.0914,
      "step": 4258
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.13982446491718292,
      "learning_rate": 1.674284018969342e-05,
      "loss": 1.1824,
      "step": 4259
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.13471931219100952,
      "learning_rate": 1.6708335769417827e-05,
      "loss": 1.1247,
      "step": 4260
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.11733231693506241,
      "learning_rate": 1.667386369866484e-05,
      "loss": 1.198,
      "step": 4261
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.12319763749837875,
      "learning_rate": 1.6639423990823012e-05,
      "loss": 1.1105,
      "step": 4262
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1368958055973053,
      "learning_rate": 1.660501665926838e-05,
      "loss": 1.1956,
      "step": 4263
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.12594982981681824,
      "learning_rate": 1.6570641717364277e-05,
      "loss": 1.1608,
      "step": 4264
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.12829987704753876,
      "learning_rate": 1.653629917846159e-05,
      "loss": 1.1041,
      "step": 4265
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.11783722043037415,
      "learning_rate": 1.6501989055898535e-05,
      "loss": 1.138,
      "step": 4266
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.15398292243480682,
      "learning_rate": 1.6467711363000793e-05,
      "loss": 1.1036,
      "step": 4267
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.12020140886306763,
      "learning_rate": 1.643346611308144e-05,
      "loss": 1.1543,
      "step": 4268
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.17364363372325897,
      "learning_rate": 1.6399253319440887e-05,
      "loss": 1.1198,
      "step": 4269
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.12527348101139069,
      "learning_rate": 1.6365072995367004e-05,
      "loss": 1.1308,
      "step": 4270
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.15316328406333923,
      "learning_rate": 1.6330925154135058e-05,
      "loss": 1.1213,
      "step": 4271
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1141105443239212,
      "learning_rate": 1.629680980900765e-05,
      "loss": 1.1944,
      "step": 4272
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1646832525730133,
      "learning_rate": 1.6262726973234843e-05,
      "loss": 1.129,
      "step": 4273
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.09889935702085495,
      "learning_rate": 1.622867666005393e-05,
      "loss": 1.2016,
      "step": 4274
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1454545557498932,
      "learning_rate": 1.619465888268972e-05,
      "loss": 1.1609,
      "step": 4275
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.13508042693138123,
      "learning_rate": 1.616067365435433e-05,
      "loss": 1.1223,
      "step": 4276
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1338137686252594,
      "learning_rate": 1.6126720988247167e-05,
      "loss": 1.0832,
      "step": 4277
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.12341427803039551,
      "learning_rate": 1.609280089755515e-05,
      "loss": 1.1363,
      "step": 4278
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.12645597755908966,
      "learning_rate": 1.605891339545237e-05,
      "loss": 1.1508,
      "step": 4279
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.15361765027046204,
      "learning_rate": 1.6025058495100385e-05,
      "loss": 1.1894,
      "step": 4280
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.13542583584785461,
      "learning_rate": 1.5991236209648053e-05,
      "loss": 1.1707,
      "step": 4281
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.15092919766902924,
      "learning_rate": 1.5957446552231526e-05,
      "loss": 1.1476,
      "step": 4282
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.14639724791049957,
      "learning_rate": 1.5923689535974305e-05,
      "loss": 1.1321,
      "step": 4283
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.15137776732444763,
      "learning_rate": 1.588996517398731e-05,
      "loss": 1.0735,
      "step": 4284
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.16955676674842834,
      "learning_rate": 1.5856273479368612e-05,
      "loss": 1.1178,
      "step": 4285
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1414966583251953,
      "learning_rate": 1.582261446520371e-05,
      "loss": 1.1734,
      "step": 4286
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.20855814218521118,
      "learning_rate": 1.5788988144565398e-05,
      "loss": 1.1546,
      "step": 4287
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1309158056974411,
      "learning_rate": 1.575539453051369e-05,
      "loss": 1.0863,
      "step": 4288
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1101151630282402,
      "learning_rate": 1.5721833636096027e-05,
      "loss": 1.068,
      "step": 4289
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.15507231652736664,
      "learning_rate": 1.568830547434703e-05,
      "loss": 1.1359,
      "step": 4290
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1428641974925995,
      "learning_rate": 1.565481005828866e-05,
      "loss": 1.0631,
      "step": 4291
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.15574949979782104,
      "learning_rate": 1.5621347400930175e-05,
      "loss": 1.1009,
      "step": 4292
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12249840050935745,
      "learning_rate": 1.5587917515268048e-05,
      "loss": 1.0866,
      "step": 4293
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12777575850486755,
      "learning_rate": 1.5554520414286067e-05,
      "loss": 1.1542,
      "step": 4294
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13583964109420776,
      "learning_rate": 1.552115611095529e-05,
      "loss": 1.1762,
      "step": 4295
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12447899580001831,
      "learning_rate": 1.5487824618234047e-05,
      "loss": 1.1012,
      "step": 4296
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.11967787146568298,
      "learning_rate": 1.54545259490679e-05,
      "loss": 1.0847,
      "step": 4297
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.11228623241186142,
      "learning_rate": 1.5421260116389637e-05,
      "loss": 1.1656,
      "step": 4298
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1130855455994606,
      "learning_rate": 1.5388027133119342e-05,
      "loss": 1.1565,
      "step": 4299
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.15877504646778107,
      "learning_rate": 1.535482701216433e-05,
      "loss": 1.107,
      "step": 4300
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13129723072052002,
      "learning_rate": 1.5321659766419128e-05,
      "loss": 1.1463,
      "step": 4301
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.14606115221977234,
      "learning_rate": 1.5288525408765564e-05,
      "loss": 1.102,
      "step": 4302
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12587065994739532,
      "learning_rate": 1.5255423952072567e-05,
      "loss": 1.21,
      "step": 4303
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1477944403886795,
      "learning_rate": 1.5222355409196398e-05,
      "loss": 1.1171,
      "step": 4304
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.17307965457439423,
      "learning_rate": 1.5189319792980516e-05,
      "loss": 1.1268,
      "step": 4305
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13188789784908295,
      "learning_rate": 1.5156317116255513e-05,
      "loss": 1.0763,
      "step": 4306
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12630468606948853,
      "learning_rate": 1.5123347391839305e-05,
      "loss": 1.0863,
      "step": 4307
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.14299114048480988,
      "learning_rate": 1.5090410632536967e-05,
      "loss": 1.1183,
      "step": 4308
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.15657897293567657,
      "learning_rate": 1.5057506851140702e-05,
      "loss": 1.0857,
      "step": 4309
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1189337819814682,
      "learning_rate": 1.5024636060429997e-05,
      "loss": 1.0612,
      "step": 4310
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13135312497615814,
      "learning_rate": 1.4991798273171465e-05,
      "loss": 1.1518,
      "step": 4311
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.18624189496040344,
      "learning_rate": 1.4958993502118901e-05,
      "loss": 1.1174,
      "step": 4312
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.11426291614770889,
      "learning_rate": 1.4926221760013392e-05,
      "loss": 1.1994,
      "step": 4313
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13417385518550873,
      "learning_rate": 1.4893483059583014e-05,
      "loss": 1.1721,
      "step": 4314
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12269094586372375,
      "learning_rate": 1.4860777413543137e-05,
      "loss": 1.1765,
      "step": 4315
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.11650824546813965,
      "learning_rate": 1.4828104834596268e-05,
      "loss": 1.137,
      "step": 4316
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.14350825548171997,
      "learning_rate": 1.4795465335432035e-05,
      "loss": 1.1657,
      "step": 4317
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.14910897612571716,
      "learning_rate": 1.4762858928727241e-05,
      "loss": 1.176,
      "step": 4318
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12682385742664337,
      "learning_rate": 1.4730285627145856e-05,
      "loss": 1.0957,
      "step": 4319
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.15082937479019165,
      "learning_rate": 1.4697745443338984e-05,
      "loss": 1.1135,
      "step": 4320
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1714775115251541,
      "learning_rate": 1.4665238389944857e-05,
      "loss": 1.1806,
      "step": 4321
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12918439507484436,
      "learning_rate": 1.4632764479588801e-05,
      "loss": 1.0697,
      "step": 4322
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13939231634140015,
      "learning_rate": 1.4600323724883335e-05,
      "loss": 1.0987,
      "step": 4323
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13208384811878204,
      "learning_rate": 1.4567916138428072e-05,
      "loss": 1.1022,
      "step": 4324
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12761560082435608,
      "learning_rate": 1.4535541732809754e-05,
      "loss": 1.1889,
      "step": 4325
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12140481919050217,
      "learning_rate": 1.4503200520602245e-05,
      "loss": 1.1743,
      "step": 4326
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.17115075886249542,
      "learning_rate": 1.4470892514366441e-05,
      "loss": 1.0801,
      "step": 4327
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1463620960712433,
      "learning_rate": 1.4438617726650439e-05,
      "loss": 1.1235,
      "step": 4328
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13678859174251556,
      "learning_rate": 1.4406376169989388e-05,
      "loss": 1.1593,
      "step": 4329
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.1374078243970871,
      "learning_rate": 1.437416785690554e-05,
      "loss": 1.1217,
      "step": 4330
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.098949134349823,
      "learning_rate": 1.4341992799908255e-05,
      "loss": 1.0738,
      "step": 4331
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13650840520858765,
      "learning_rate": 1.4309851011493903e-05,
      "loss": 1.1523,
      "step": 4332
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.14365199208259583,
      "learning_rate": 1.4277742504146008e-05,
      "loss": 1.0738,
      "step": 4333
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12990158796310425,
      "learning_rate": 1.4245667290335174e-05,
      "loss": 1.1272,
      "step": 4334
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.14570455253124237,
      "learning_rate": 1.4213625382518968e-05,
      "loss": 1.0701,
      "step": 4335
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12693321704864502,
      "learning_rate": 1.4181616793142172e-05,
      "loss": 1.1613,
      "step": 4336
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.09348238259553909,
      "learning_rate": 1.4149641534636549e-05,
      "loss": 1.0948,
      "step": 4337
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12044403702020645,
      "learning_rate": 1.4117699619420877e-05,
      "loss": 1.0929,
      "step": 4338
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.13622894883155823,
      "learning_rate": 1.4085791059901076e-05,
      "loss": 1.0468,
      "step": 4339
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.16004148125648499,
      "learning_rate": 1.4053915868470013e-05,
      "loss": 1.0834,
      "step": 4340
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1384895294904709,
      "learning_rate": 1.402207405750765e-05,
      "loss": 1.0803,
      "step": 4341
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15442202985286713,
      "learning_rate": 1.399026563938105e-05,
      "loss": 1.1165,
      "step": 4342
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.14327654242515564,
      "learning_rate": 1.3958490626444154e-05,
      "loss": 1.1475,
      "step": 4343
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13714727759361267,
      "learning_rate": 1.3926749031038056e-05,
      "loss": 1.0492,
      "step": 4344
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13002392649650574,
      "learning_rate": 1.3895040865490816e-05,
      "loss": 1.1035,
      "step": 4345
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13084842264652252,
      "learning_rate": 1.3863366142117506e-05,
      "loss": 1.1015,
      "step": 4346
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15821552276611328,
      "learning_rate": 1.383172487322023e-05,
      "loss": 1.0772,
      "step": 4347
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13527686893939972,
      "learning_rate": 1.3800117071088104e-05,
      "loss": 1.1348,
      "step": 4348
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13305167853832245,
      "learning_rate": 1.3768542747997215e-05,
      "loss": 1.0837,
      "step": 4349
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.14797480404376984,
      "learning_rate": 1.3737001916210713e-05,
      "loss": 1.1431,
      "step": 4350
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.10989313572645187,
      "learning_rate": 1.3705494587978628e-05,
      "loss": 1.141,
      "step": 4351
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.12826380133628845,
      "learning_rate": 1.3674020775538077e-05,
      "loss": 1.1173,
      "step": 4352
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13416388630867004,
      "learning_rate": 1.3642580491113122e-05,
      "loss": 1.0883,
      "step": 4353
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15589915215969086,
      "learning_rate": 1.3611173746914795e-05,
      "loss": 1.1375,
      "step": 4354
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15658612549304962,
      "learning_rate": 1.3579800555141165e-05,
      "loss": 1.0553,
      "step": 4355
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.14531481266021729,
      "learning_rate": 1.3548460927977159e-05,
      "loss": 1.1361,
      "step": 4356
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.11375796049833298,
      "learning_rate": 1.351715487759474e-05,
      "loss": 1.0644,
      "step": 4357
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.14637741446495056,
      "learning_rate": 1.3485882416152818e-05,
      "loss": 1.1221,
      "step": 4358
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.20275047421455383,
      "learning_rate": 1.3454643555797274e-05,
      "loss": 1.1481,
      "step": 4359
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15003275871276855,
      "learning_rate": 1.342343830866093e-05,
      "loss": 1.1257,
      "step": 4360
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1290082484483719,
      "learning_rate": 1.3392266686863509e-05,
      "loss": 1.1208,
      "step": 4361
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.19464193284511566,
      "learning_rate": 1.3361128702511716e-05,
      "loss": 1.1326,
      "step": 4362
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15163567662239075,
      "learning_rate": 1.3330024367699223e-05,
      "loss": 1.1521,
      "step": 4363
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.11787136644124985,
      "learning_rate": 1.3298953694506522e-05,
      "loss": 1.117,
      "step": 4364
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13570216298103333,
      "learning_rate": 1.326791669500117e-05,
      "loss": 1.0739,
      "step": 4365
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13937844336032867,
      "learning_rate": 1.3236913381237592e-05,
      "loss": 1.0998,
      "step": 4366
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.11584049463272095,
      "learning_rate": 1.3205943765257055e-05,
      "loss": 1.0762,
      "step": 4367
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1812932938337326,
      "learning_rate": 1.317500785908783e-05,
      "loss": 1.0696,
      "step": 4368
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15424339473247528,
      "learning_rate": 1.31441056747451e-05,
      "loss": 1.0852,
      "step": 4369
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.18165987730026245,
      "learning_rate": 1.3113237224230834e-05,
      "loss": 1.1657,
      "step": 4370
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.12156308442354202,
      "learning_rate": 1.3082402519534076e-05,
      "loss": 1.1529,
      "step": 4371
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15393023192882538,
      "learning_rate": 1.3051601572630612e-05,
      "loss": 1.1008,
      "step": 4372
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.18390028178691864,
      "learning_rate": 1.3020834395483195e-05,
      "loss": 1.1784,
      "step": 4373
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.11997378617525101,
      "learning_rate": 1.2990101000041444e-05,
      "loss": 1.0865,
      "step": 4374
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.14333024621009827,
      "learning_rate": 1.2959401398241843e-05,
      "loss": 1.0836,
      "step": 4375
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.10710073262453079,
      "learning_rate": 1.2928735602007769e-05,
      "loss": 1.1522,
      "step": 4376
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15028801560401917,
      "learning_rate": 1.2898103623249457e-05,
      "loss": 1.1144,
      "step": 4377
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13241040706634521,
      "learning_rate": 1.2867505473864028e-05,
      "loss": 1.0599,
      "step": 4378
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15825334191322327,
      "learning_rate": 1.2836941165735462e-05,
      "loss": 1.1245,
      "step": 4379
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.10357356816530228,
      "learning_rate": 1.280641071073455e-05,
      "loss": 1.0995,
      "step": 4380
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13159774243831635,
      "learning_rate": 1.277591412071899e-05,
      "loss": 1.1092,
      "step": 4381
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.16397859156131744,
      "learning_rate": 1.2745451407533293e-05,
      "loss": 1.0542,
      "step": 4382
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.18642419576644897,
      "learning_rate": 1.271502258300885e-05,
      "loss": 1.1317,
      "step": 4383
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.15759436786174774,
      "learning_rate": 1.2684627658963867e-05,
      "loss": 1.0571,
      "step": 4384
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.13173717260360718,
      "learning_rate": 1.2654266647203339e-05,
      "loss": 1.1821,
      "step": 4385
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1855558454990387,
      "learning_rate": 1.262393955951916e-05,
      "loss": 1.1442,
      "step": 4386
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.12076414376497269,
      "learning_rate": 1.2593646407690051e-05,
      "loss": 1.0724,
      "step": 4387
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1347951740026474,
      "learning_rate": 1.2563387203481448e-05,
      "loss": 1.0831,
      "step": 4388
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1579679697751999,
      "learning_rate": 1.2533161958645756e-05,
      "loss": 1.1586,
      "step": 4389
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.14599944651126862,
      "learning_rate": 1.2502970684922066e-05,
      "loss": 1.0941,
      "step": 4390
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.10327880084514618,
      "learning_rate": 1.2472813394036343e-05,
      "loss": 1.0697,
      "step": 4391
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1522255688905716,
      "learning_rate": 1.2442690097701327e-05,
      "loss": 1.165,
      "step": 4392
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.11582104116678238,
      "learning_rate": 1.2412600807616525e-05,
      "loss": 1.0918,
      "step": 4393
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.14551933109760284,
      "learning_rate": 1.2382545535468315e-05,
      "loss": 1.0726,
      "step": 4394
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.13929572701454163,
      "learning_rate": 1.2352524292929824e-05,
      "loss": 1.1189,
      "step": 4395
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.14231844246387482,
      "learning_rate": 1.2322537091660912e-05,
      "loss": 1.108,
      "step": 4396
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.10934177786111832,
      "learning_rate": 1.22925839433083e-05,
      "loss": 1.067,
      "step": 4397
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.13370361924171448,
      "learning_rate": 1.2262664859505435e-05,
      "loss": 1.1001,
      "step": 4398
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.15218055248260498,
      "learning_rate": 1.223277985187251e-05,
      "loss": 1.2108,
      "step": 4399
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.19181691110134125,
      "learning_rate": 1.2202928932016589e-05,
      "loss": 1.0684,
      "step": 4400
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.16071465611457825,
      "learning_rate": 1.217311211153137e-05,
      "loss": 1.063,
      "step": 4401
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.16756823658943176,
      "learning_rate": 1.2143329401997372e-05,
      "loss": 1.1832,
      "step": 4402
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.16241203248500824,
      "learning_rate": 1.2113580814981884e-05,
      "loss": 1.0531,
      "step": 4403
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1533583402633667,
      "learning_rate": 1.2083866362038865e-05,
      "loss": 1.1533,
      "step": 4404
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1352750062942505,
      "learning_rate": 1.2054186054709105e-05,
      "loss": 1.1317,
      "step": 4405
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.10988578200340271,
      "learning_rate": 1.2024539904520072e-05,
      "loss": 1.1047,
      "step": 4406
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.12754975259304047,
      "learning_rate": 1.1994927922985998e-05,
      "loss": 1.1701,
      "step": 4407
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.12251193821430206,
      "learning_rate": 1.1965350121607865e-05,
      "loss": 1.1464,
      "step": 4408
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.10449106991291046,
      "learning_rate": 1.1935806511873304e-05,
      "loss": 1.0633,
      "step": 4409
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.13005949556827545,
      "learning_rate": 1.1906297105256725e-05,
      "loss": 1.1778,
      "step": 4410
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.10989183932542801,
      "learning_rate": 1.187682191321925e-05,
      "loss": 1.0804,
      "step": 4411
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.16075707972049713,
      "learning_rate": 1.1847380947208698e-05,
      "loss": 1.2107,
      "step": 4412
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.16259409487247467,
      "learning_rate": 1.1817974218659622e-05,
      "loss": 1.1293,
      "step": 4413
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.14562450349330902,
      "learning_rate": 1.178860173899321e-05,
      "loss": 1.1005,
      "step": 4414
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1152070015668869,
      "learning_rate": 1.1759263519617437e-05,
      "loss": 1.2603,
      "step": 4415
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.12120033800601959,
      "learning_rate": 1.172995957192693e-05,
      "loss": 1.1266,
      "step": 4416
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.14844419062137604,
      "learning_rate": 1.1700689907302953e-05,
      "loss": 1.1415,
      "step": 4417
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1310236155986786,
      "learning_rate": 1.167145453711358e-05,
      "loss": 1.2137,
      "step": 4418
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.11370284110307693,
      "learning_rate": 1.1642253472713426e-05,
      "loss": 1.1368,
      "step": 4419
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1356082707643509,
      "learning_rate": 1.161308672544389e-05,
      "loss": 1.2084,
      "step": 4420
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.15232105553150177,
      "learning_rate": 1.1583954306633005e-05,
      "loss": 1.1407,
      "step": 4421
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1689256876707077,
      "learning_rate": 1.1554856227595433e-05,
      "loss": 1.0842,
      "step": 4422
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1273879110813141,
      "learning_rate": 1.1525792499632527e-05,
      "loss": 1.2418,
      "step": 4423
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1273879110813141,
      "learning_rate": 1.1525792499632527e-05,
      "loss": 1.0938,
      "step": 4424
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.13114044070243835,
      "learning_rate": 1.1496763134032363e-05,
      "loss": 1.0476,
      "step": 4425
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.12532052397727966,
      "learning_rate": 1.1467768142069546e-05,
      "loss": 1.0752,
      "step": 4426
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.13807563483715057,
      "learning_rate": 1.1438807535005436e-05,
      "loss": 1.1922,
      "step": 4427
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1879110336303711,
      "learning_rate": 1.1409881324088011e-05,
      "loss": 1.169,
      "step": 4428
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.14134246110916138,
      "learning_rate": 1.138098952055181e-05,
      "loss": 1.2198,
      "step": 4429
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.11935198307037354,
      "learning_rate": 1.1352132135618166e-05,
      "loss": 1.186,
      "step": 4430
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.19223986566066742,
      "learning_rate": 1.1323309180494901e-05,
      "loss": 1.1436,
      "step": 4431
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.15105709433555603,
      "learning_rate": 1.1294520666376518e-05,
      "loss": 1.1086,
      "step": 4432
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.117209792137146,
      "learning_rate": 1.1265766604444172e-05,
      "loss": 1.0543,
      "step": 4433
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.12545882165431976,
      "learning_rate": 1.1237047005865576e-05,
      "loss": 1.1604,
      "step": 4434
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.14809970557689667,
      "learning_rate": 1.1208361881795116e-05,
      "loss": 1.1631,
      "step": 4435
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1574985235929489,
      "learning_rate": 1.1179711243373736e-05,
      "loss": 1.1267,
      "step": 4436
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.14643482863903046,
      "learning_rate": 1.1151095101729047e-05,
      "loss": 1.1347,
      "step": 4437
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1628008931875229,
      "learning_rate": 1.1122513467975237e-05,
      "loss": 1.099,
      "step": 4438
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1239699274301529,
      "learning_rate": 1.1093966353213037e-05,
      "loss": 1.0106,
      "step": 4439
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.11441464722156525,
      "learning_rate": 1.1065453768529843e-05,
      "loss": 1.1082,
      "step": 4440
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1353859305381775,
      "learning_rate": 1.1036975724999609e-05,
      "loss": 1.1588,
      "step": 4441
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.11166424304246902,
      "learning_rate": 1.1008532233682877e-05,
      "loss": 1.1071,
      "step": 4442
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.12025504559278488,
      "learning_rate": 1.098012330562681e-05,
      "loss": 1.1082,
      "step": 4443
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.12716332077980042,
      "learning_rate": 1.0951748951865048e-05,
      "loss": 1.1338,
      "step": 4444
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.134698286652565,
      "learning_rate": 1.0923409183417887e-05,
      "loss": 1.1613,
      "step": 4445
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1559639573097229,
      "learning_rate": 1.08951040112922e-05,
      "loss": 1.1309,
      "step": 4446
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.13860522210597992,
      "learning_rate": 1.0866833446481317e-05,
      "loss": 1.0728,
      "step": 4447
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.14828701317310333,
      "learning_rate": 1.0838597499965275e-05,
      "loss": 1.1616,
      "step": 4448
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1638706773519516,
      "learning_rate": 1.0810396182710536e-05,
      "loss": 1.0597,
      "step": 4449
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.12956419587135315,
      "learning_rate": 1.0782229505670193e-05,
      "loss": 1.1088,
      "step": 4450
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.13180987536907196,
      "learning_rate": 1.0754097479783876e-05,
      "loss": 1.108,
      "step": 4451
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.17648892104625702,
      "learning_rate": 1.0726000115977696e-05,
      "loss": 1.0803,
      "step": 4452
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.13340191543102264,
      "learning_rate": 1.069793742516435e-05,
      "loss": 1.1676,
      "step": 4453
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1255689412355423,
      "learning_rate": 1.0669909418243118e-05,
      "loss": 1.1307,
      "step": 4454
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1431410312652588,
      "learning_rate": 1.0641916106099692e-05,
      "loss": 1.2167,
      "step": 4455
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.12132400274276733,
      "learning_rate": 1.0613957499606386e-05,
      "loss": 1.0425,
      "step": 4456
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.16257156431674957,
      "learning_rate": 1.0586033609622004e-05,
      "loss": 1.0947,
      "step": 4457
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.15706345438957214,
      "learning_rate": 1.0558144446991836e-05,
      "loss": 1.1558,
      "step": 4458
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.12804453074932098,
      "learning_rate": 1.0530290022547728e-05,
      "loss": 1.1164,
      "step": 4459
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.12935525178909302,
      "learning_rate": 1.0502470347108018e-05,
      "loss": 1.1046,
      "step": 4460
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.15077902376651764,
      "learning_rate": 1.0474685431477539e-05,
      "loss": 1.137,
      "step": 4461
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.11985961347818375,
      "learning_rate": 1.0446935286447656e-05,
      "loss": 1.1563,
      "step": 4462
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1570107489824295,
      "learning_rate": 1.0419219922796176e-05,
      "loss": 1.2238,
      "step": 4463
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.16242259740829468,
      "learning_rate": 1.0391539351287439e-05,
      "loss": 1.1296,
      "step": 4464
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1268274039030075,
      "learning_rate": 1.0363893582672247e-05,
      "loss": 1.0995,
      "step": 4465
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.14929479360580444,
      "learning_rate": 1.0336282627687921e-05,
      "loss": 1.1229,
      "step": 4466
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1363082379102707,
      "learning_rate": 1.030870649705825e-05,
      "loss": 1.1501,
      "step": 4467
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1720314770936966,
      "learning_rate": 1.0281165201493438e-05,
      "loss": 1.0902,
      "step": 4468
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1120753139257431,
      "learning_rate": 1.025365875169023e-05,
      "loss": 1.1416,
      "step": 4469
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1296852082014084,
      "learning_rate": 1.0226187158331823e-05,
      "loss": 1.163,
      "step": 4470
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.12862052023410797,
      "learning_rate": 1.0198750432087855e-05,
      "loss": 1.1244,
      "step": 4471
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.11146964877843857,
      "learning_rate": 1.0171348583614459e-05,
      "loss": 1.1607,
      "step": 4472
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.16124053299427032,
      "learning_rate": 1.0143981623554156e-05,
      "loss": 1.1624,
      "step": 4473
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.12639330327510834,
      "learning_rate": 1.0116649562535985e-05,
      "loss": 1.0322,
      "step": 4474
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.11445170640945435,
      "learning_rate": 1.0089352411175423e-05,
      "loss": 1.0626,
      "step": 4475
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.14717185497283936,
      "learning_rate": 1.0062090180074291e-05,
      "loss": 1.0987,
      "step": 4476
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.16013987362384796,
      "learning_rate": 1.0034862879821027e-05,
      "loss": 1.0469,
      "step": 4477
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.14157843589782715,
      "learning_rate": 1.000767052099033e-05,
      "loss": 1.1201,
      "step": 4478
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.14818736910820007,
      "learning_rate": 9.98051311414342e-06,
      "loss": 1.1229,
      "step": 4479
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.15598857402801514,
      "learning_rate": 9.953390669827945e-06,
      "loss": 1.0745,
      "step": 4480
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.12923315167427063,
      "learning_rate": 9.926303198577913e-06,
      "loss": 1.0871,
      "step": 4481
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.11020199954509735,
      "learning_rate": 9.899250710913765e-06,
      "loss": 1.1439,
      "step": 4482
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.16335777938365936,
      "learning_rate": 9.872233217342463e-06,
      "loss": 1.1294,
      "step": 4483
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1412651538848877,
      "learning_rate": 9.845250728357215e-06,
      "loss": 1.1725,
      "step": 4484
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.16336454451084137,
      "learning_rate": 9.818303254437722e-06,
      "loss": 1.0979,
      "step": 4485
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.14288246631622314,
      "learning_rate": 9.791390806050115e-06,
      "loss": 1.1256,
      "step": 4486
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1322931945323944,
      "learning_rate": 9.764513393646812e-06,
      "loss": 1.1048,
      "step": 4487
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.12442046403884888,
      "learning_rate": 9.737671027666728e-06,
      "loss": 1.1621,
      "step": 4488
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.13627880811691284,
      "learning_rate": 9.710863718535135e-06,
      "loss": 1.1526,
      "step": 4489
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1280861645936966,
      "learning_rate": 9.684091476663659e-06,
      "loss": 1.1011,
      "step": 4490
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.14453792572021484,
      "learning_rate": 9.657354312450362e-06,
      "loss": 1.1696,
      "step": 4491
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.15191799402236938,
      "learning_rate": 9.630652236279625e-06,
      "loss": 1.1368,
      "step": 4492
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.13446149230003357,
      "learning_rate": 9.603985258522218e-06,
      "loss": 1.0745,
      "step": 4493
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.15580956637859344,
      "learning_rate": 9.577353389535315e-06,
      "loss": 1.1085,
      "step": 4494
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.11491867899894714,
      "learning_rate": 9.550756639662417e-06,
      "loss": 1.1481,
      "step": 4495
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.13143910467624664,
      "learning_rate": 9.524195019233407e-06,
      "loss": 1.2259,
      "step": 4496
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.13902612030506134,
      "learning_rate": 9.497668538564474e-06,
      "loss": 1.0773,
      "step": 4497
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.11181966215372086,
      "learning_rate": 9.471177207958237e-06,
      "loss": 1.0505,
      "step": 4498
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1706608384847641,
      "learning_rate": 9.444721037703597e-06,
      "loss": 1.2644,
      "step": 4499
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14175070822238922,
      "learning_rate": 9.418300038075845e-06,
      "loss": 1.0227,
      "step": 4500
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12038809806108475,
      "learning_rate": 9.391914219336606e-06,
      "loss": 1.1372,
      "step": 4501
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.11803561449050903,
      "learning_rate": 9.365563591733784e-06,
      "loss": 1.16,
      "step": 4502
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.15593959391117096,
      "learning_rate": 9.33924816550168e-06,
      "loss": 1.0438,
      "step": 4503
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.16889461874961853,
      "learning_rate": 9.31296795086093e-06,
      "loss": 1.1784,
      "step": 4504
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12889009714126587,
      "learning_rate": 9.28672295801839e-06,
      "loss": 1.1716,
      "step": 4505
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12882664799690247,
      "learning_rate": 9.260513197167397e-06,
      "loss": 1.2184,
      "step": 4506
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14612983167171478,
      "learning_rate": 9.234338678487509e-06,
      "loss": 1.169,
      "step": 4507
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12072368711233139,
      "learning_rate": 9.208199412144558e-06,
      "loss": 1.083,
      "step": 4508
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12626969814300537,
      "learning_rate": 9.182095408290781e-06,
      "loss": 1.1443,
      "step": 4509
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.10508429259061813,
      "learning_rate": 9.156026677064632e-06,
      "loss": 1.0122,
      "step": 4510
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1407318115234375,
      "learning_rate": 9.129993228590916e-06,
      "loss": 1.1645,
      "step": 4511
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12766827642917633,
      "learning_rate": 9.103995072980765e-06,
      "loss": 1.1298,
      "step": 4512
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12230279296636581,
      "learning_rate": 9.078032220331522e-06,
      "loss": 1.0974,
      "step": 4513
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12815696001052856,
      "learning_rate": 9.052104680726858e-06,
      "loss": 1.1487,
      "step": 4514
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14419355988502502,
      "learning_rate": 9.026212464236772e-06,
      "loss": 1.1865,
      "step": 4515
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1427794098854065,
      "learning_rate": 9.000355580917463e-06,
      "loss": 1.0792,
      "step": 4516
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.16184915602207184,
      "learning_rate": 8.974534040811445e-06,
      "loss": 1.1204,
      "step": 4517
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14943468570709229,
      "learning_rate": 8.948747853947526e-06,
      "loss": 1.1231,
      "step": 4518
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14052538573741913,
      "learning_rate": 8.922997030340752e-06,
      "loss": 1.1238,
      "step": 4519
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14285975694656372,
      "learning_rate": 8.897281579992467e-06,
      "loss": 0.9974,
      "step": 4520
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14195679128170013,
      "learning_rate": 8.871601512890238e-06,
      "loss": 1.1639,
      "step": 4521
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14225901663303375,
      "learning_rate": 8.845956839007896e-06,
      "loss": 1.1014,
      "step": 4522
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12923932075500488,
      "learning_rate": 8.820347568305542e-06,
      "loss": 1.0422,
      "step": 4523
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.11818478256464005,
      "learning_rate": 8.794773710729542e-06,
      "loss": 1.0289,
      "step": 4524
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12879891693592072,
      "learning_rate": 8.769235276212495e-06,
      "loss": 1.1471,
      "step": 4525
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1382279396057129,
      "learning_rate": 8.74373227467319e-06,
      "loss": 1.1305,
      "step": 4526
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12976516783237457,
      "learning_rate": 8.718264716016722e-06,
      "loss": 1.0007,
      "step": 4527
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.11358962208032608,
      "learning_rate": 8.692832610134427e-06,
      "loss": 1.0987,
      "step": 4528
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.10135991871356964,
      "learning_rate": 8.66743596690377e-06,
      "loss": 1.1171,
      "step": 4529
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1490253359079361,
      "learning_rate": 8.642074796188592e-06,
      "loss": 1.1059,
      "step": 4530
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14218680560588837,
      "learning_rate": 8.61674910783884e-06,
      "loss": 1.1361,
      "step": 4531
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12580186128616333,
      "learning_rate": 8.59145891169072e-06,
      "loss": 1.1319,
      "step": 4532
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1835440844297409,
      "learning_rate": 8.566204217566664e-06,
      "loss": 1.0987,
      "step": 4533
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.14689713716506958,
      "learning_rate": 8.540985035275272e-06,
      "loss": 1.0142,
      "step": 4534
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1454998105764389,
      "learning_rate": 8.515801374611432e-06,
      "loss": 1.0977,
      "step": 4535
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.16207049787044525,
      "learning_rate": 8.490653245356185e-06,
      "loss": 1.1983,
      "step": 4536
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.11399718374013901,
      "learning_rate": 8.465540657276727e-06,
      "loss": 1.0691,
      "step": 4537
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.21277551352977753,
      "learning_rate": 8.44046362012656e-06,
      "loss": 1.0895,
      "step": 4538
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.11614827811717987,
      "learning_rate": 8.415422143645245e-06,
      "loss": 1.1394,
      "step": 4539
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.13415004312992096,
      "learning_rate": 8.39041623755864e-06,
      "loss": 1.2008,
      "step": 4540
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2049035280942917,
      "learning_rate": 8.365445911578785e-06,
      "loss": 1.0262,
      "step": 4541
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12968619167804718,
      "learning_rate": 8.340511175403809e-06,
      "loss": 1.0541,
      "step": 4542
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1305578052997589,
      "learning_rate": 8.315612038718101e-06,
      "loss": 0.9867,
      "step": 4543
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1413247138261795,
      "learning_rate": 8.290748511192215e-06,
      "loss": 1.1438,
      "step": 4544
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12361438572406769,
      "learning_rate": 8.265920602482824e-06,
      "loss": 1.0811,
      "step": 4545
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.173883318901062,
      "learning_rate": 8.241128322232817e-06,
      "loss": 1.1734,
      "step": 4546
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.11956268548965454,
      "learning_rate": 8.216371680071244e-06,
      "loss": 1.0777,
      "step": 4547
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.12703682482242584,
      "learning_rate": 8.191650685613272e-06,
      "loss": 1.1503,
      "step": 4548
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11919744312763214,
      "learning_rate": 8.166965348460299e-06,
      "loss": 1.1301,
      "step": 4549
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.16681894659996033,
      "learning_rate": 8.142315678199763e-06,
      "loss": 1.1781,
      "step": 4550
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.10657525062561035,
      "learning_rate": 8.117701684405343e-06,
      "loss": 1.0561,
      "step": 4551
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.15076905488967896,
      "learning_rate": 8.093123376636836e-06,
      "loss": 1.0688,
      "step": 4552
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.14618414640426636,
      "learning_rate": 8.06858076444017e-06,
      "loss": 1.0854,
      "step": 4553
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1816091686487198,
      "learning_rate": 8.044073857347423e-06,
      "loss": 1.1273,
      "step": 4554
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11911457777023315,
      "learning_rate": 8.019602664876758e-06,
      "loss": 1.0384,
      "step": 4555
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.14835257828235626,
      "learning_rate": 7.995167196532527e-06,
      "loss": 1.1391,
      "step": 4556
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11910203099250793,
      "learning_rate": 7.970767461805217e-06,
      "loss": 1.1131,
      "step": 4557
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.157576322555542,
      "learning_rate": 7.946403470171327e-06,
      "loss": 1.0939,
      "step": 4558
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.14588125050067902,
      "learning_rate": 7.922075231093629e-06,
      "loss": 1.0642,
      "step": 4559
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.10849962383508682,
      "learning_rate": 7.897782754020888e-06,
      "loss": 1.1041,
      "step": 4560
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11644388735294342,
      "learning_rate": 7.873526048388026e-06,
      "loss": 1.1635,
      "step": 4561
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1264728605747223,
      "learning_rate": 7.849305123616092e-06,
      "loss": 1.0855,
      "step": 4562
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.18656310439109802,
      "learning_rate": 7.825119989112173e-06,
      "loss": 1.1245,
      "step": 4563
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.12174135446548462,
      "learning_rate": 7.800970654269512e-06,
      "loss": 1.0802,
      "step": 4564
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11160901933908463,
      "learning_rate": 7.776857128467463e-06,
      "loss": 1.1221,
      "step": 4565
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11971074342727661,
      "learning_rate": 7.75277942107141e-06,
      "loss": 1.2209,
      "step": 4566
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.13110855221748352,
      "learning_rate": 7.728737541432863e-06,
      "loss": 1.0532,
      "step": 4567
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.126364067196846,
      "learning_rate": 7.704731498889428e-06,
      "loss": 1.0007,
      "step": 4568
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.12909811735153198,
      "learning_rate": 7.680761302764727e-06,
      "loss": 1.0721,
      "step": 4569
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.13461457192897797,
      "learning_rate": 7.656826962368579e-06,
      "loss": 1.1037,
      "step": 4570
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.15062595903873444,
      "learning_rate": 7.632928486996749e-06,
      "loss": 1.1418,
      "step": 4571
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.12713488936424255,
      "learning_rate": 7.609065885931155e-06,
      "loss": 1.1683,
      "step": 4572
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.14793823659420013,
      "learning_rate": 7.585239168439761e-06,
      "loss": 1.0447,
      "step": 4573
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.10098742693662643,
      "learning_rate": 7.561448343776567e-06,
      "loss": 1.0379,
      "step": 4574
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.16674096882343292,
      "learning_rate": 7.537693421181658e-06,
      "loss": 1.1134,
      "step": 4575
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11884458363056183,
      "learning_rate": 7.513974409881186e-06,
      "loss": 1.1222,
      "step": 4576
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.12550874054431915,
      "learning_rate": 7.490291319087339e-06,
      "loss": 1.0651,
      "step": 4577
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11711330711841583,
      "learning_rate": 7.46664415799837e-06,
      "loss": 1.1201,
      "step": 4578
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.14291930198669434,
      "learning_rate": 7.443032935798533e-06,
      "loss": 1.1214,
      "step": 4579
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11966061592102051,
      "learning_rate": 7.419457661658169e-06,
      "loss": 1.0719,
      "step": 4580
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.16177070140838623,
      "learning_rate": 7.3959183447336434e-06,
      "loss": 1.0394,
      "step": 4581
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.14253464341163635,
      "learning_rate": 7.372414994167354e-06,
      "loss": 1.0812,
      "step": 4582
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.12980598211288452,
      "learning_rate": 7.3489476190877536e-06,
      "loss": 1.119,
      "step": 4583
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1438576579093933,
      "learning_rate": 7.325516228609264e-06,
      "loss": 1.1445,
      "step": 4584
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.12376336008310318,
      "learning_rate": 7.302120831832382e-06,
      "loss": 1.1555,
      "step": 4585
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1499795913696289,
      "learning_rate": 7.278761437843629e-06,
      "loss": 1.1272,
      "step": 4586
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1276843249797821,
      "learning_rate": 7.255438055715469e-06,
      "loss": 1.1199,
      "step": 4587
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.17369627952575684,
      "learning_rate": 7.232150694506512e-06,
      "loss": 1.0958,
      "step": 4588
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.12218989431858063,
      "learning_rate": 7.208899363261234e-06,
      "loss": 1.1636,
      "step": 4589
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.11500949412584305,
      "learning_rate": 7.185684071010224e-06,
      "loss": 1.1377,
      "step": 4590
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.21784934401512146,
      "learning_rate": 7.162504826770033e-06,
      "loss": 1.0881,
      "step": 4591
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.15630505979061127,
      "learning_rate": 7.139361639543185e-06,
      "loss": 1.1878,
      "step": 4592
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.13532191514968872,
      "learning_rate": 7.116254518318221e-06,
      "loss": 1.0299,
      "step": 4593
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.17679186165332794,
      "learning_rate": 7.0931834720697535e-06,
      "loss": 1.1992,
      "step": 4594
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.12471570819616318,
      "learning_rate": 7.070148509758223e-06,
      "loss": 1.1301,
      "step": 4595
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.18550701439380646,
      "learning_rate": 7.0471496403301975e-06,
      "loss": 1.1117,
      "step": 4596
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.133230522274971,
      "learning_rate": 7.024186872718164e-06,
      "loss": 1.1149,
      "step": 4597
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.14353294670581818,
      "learning_rate": 7.001260215840566e-06,
      "loss": 1.1749,
      "step": 4598
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.13788101077079773,
      "learning_rate": 6.978369678601893e-06,
      "loss": 1.1028,
      "step": 4599
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.12963920831680298,
      "learning_rate": 6.955515269892532e-06,
      "loss": 1.0604,
      "step": 4600
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.13628841936588287,
      "learning_rate": 6.932696998588895e-06,
      "loss": 1.1749,
      "step": 4601
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14475880563259125,
      "learning_rate": 6.909914873553347e-06,
      "loss": 1.1649,
      "step": 4602
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.11575454473495483,
      "learning_rate": 6.887168903634178e-06,
      "loss": 1.0757,
      "step": 4603
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14761561155319214,
      "learning_rate": 6.864459097665654e-06,
      "loss": 1.1133,
      "step": 4604
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.15299974381923676,
      "learning_rate": 6.8417854644680205e-06,
      "loss": 1.126,
      "step": 4605
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.11659429222345352,
      "learning_rate": 6.819148012847454e-06,
      "loss": 1.1009,
      "step": 4606
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1450616866350174,
      "learning_rate": 6.796546751596089e-06,
      "loss": 1.0824,
      "step": 4607
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1936507672071457,
      "learning_rate": 6.773981689491959e-06,
      "loss": 1.1258,
      "step": 4608
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.15021418035030365,
      "learning_rate": 6.75145283529911e-06,
      "loss": 1.0262,
      "step": 4609
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1538759171962738,
      "learning_rate": 6.728960197767475e-06,
      "loss": 1.095,
      "step": 4610
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14356644451618195,
      "learning_rate": 6.706503785632934e-06,
      "loss": 1.1465,
      "step": 4611
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.11980709433555603,
      "learning_rate": 6.68408360761732e-06,
      "loss": 1.1338,
      "step": 4612
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.13383324444293976,
      "learning_rate": 6.661699672428334e-06,
      "loss": 1.1391,
      "step": 4613
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14389735460281372,
      "learning_rate": 6.639351988759657e-06,
      "loss": 1.2052,
      "step": 4614
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1403246521949768,
      "learning_rate": 6.6170405652909e-06,
      "loss": 1.121,
      "step": 4615
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.15451696515083313,
      "learning_rate": 6.594765410687487e-06,
      "loss": 1.0851,
      "step": 4616
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.13450634479522705,
      "learning_rate": 6.57252653360092e-06,
      "loss": 1.1046,
      "step": 4617
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14369475841522217,
      "learning_rate": 6.550323942668468e-06,
      "loss": 1.0603,
      "step": 4618
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.15464037656784058,
      "learning_rate": 6.528157646513378e-06,
      "loss": 1.1232,
      "step": 4619
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.15372851490974426,
      "learning_rate": 6.506027653744795e-06,
      "loss": 1.0524,
      "step": 4620
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14126889407634735,
      "learning_rate": 6.483933972957734e-06,
      "loss": 1.0828,
      "step": 4621
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.10677427053451538,
      "learning_rate": 6.461876612733109e-06,
      "loss": 1.0721,
      "step": 4622
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.12105177342891693,
      "learning_rate": 6.439855581637799e-06,
      "loss": 1.1578,
      "step": 4623
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.11271104961633682,
      "learning_rate": 6.417870888224475e-06,
      "loss": 1.0932,
      "step": 4624
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1419820785522461,
      "learning_rate": 6.39592254103174e-06,
      "loss": 1.088,
      "step": 4625
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.13897907733917236,
      "learning_rate": 6.37401054858412e-06,
      "loss": 1.1166,
      "step": 4626
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.10530392825603485,
      "learning_rate": 6.352134919391928e-06,
      "loss": 1.1496,
      "step": 4627
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.12854984402656555,
      "learning_rate": 6.330295661951435e-06,
      "loss": 1.2621,
      "step": 4628
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14800739288330078,
      "learning_rate": 6.308492784744746e-06,
      "loss": 1.1298,
      "step": 4629
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.15473537147045135,
      "learning_rate": 6.2867262962398535e-06,
      "loss": 1.1157,
      "step": 4630
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.11431333422660828,
      "learning_rate": 6.2649962048906276e-06,
      "loss": 1.2743,
      "step": 4631
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1464116871356964,
      "learning_rate": 6.24330251913674e-06,
      "loss": 1.064,
      "step": 4632
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.12025155127048492,
      "learning_rate": 6.221645247403807e-06,
      "loss": 1.0853,
      "step": 4633
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1460656076669693,
      "learning_rate": 6.200024398103255e-06,
      "loss": 1.1389,
      "step": 4634
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.19698350131511688,
      "learning_rate": 6.178439979632366e-06,
      "loss": 1.0766,
      "step": 4635
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.120268814265728,
      "learning_rate": 6.156892000374292e-06,
      "loss": 1.1301,
      "step": 4636
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.10906530916690826,
      "learning_rate": 6.135380468698005e-06,
      "loss": 0.9966,
      "step": 4637
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14245474338531494,
      "learning_rate": 6.113905392958341e-06,
      "loss": 1.1713,
      "step": 4638
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.13036566972732544,
      "learning_rate": 6.092466781495976e-06,
      "loss": 1.1069,
      "step": 4639
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.10970976948738098,
      "learning_rate": 6.071064642637403e-06,
      "loss": 1.0654,
      "step": 4640
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.16147339344024658,
      "learning_rate": 6.049698984695007e-06,
      "loss": 1.2147,
      "step": 4641
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14279448986053467,
      "learning_rate": 6.028369815966917e-06,
      "loss": 1.0699,
      "step": 4642
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.13877791166305542,
      "learning_rate": 6.007077144737161e-06,
      "loss": 1.0906,
      "step": 4643
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.13291138410568237,
      "learning_rate": 5.985820979275569e-06,
      "loss": 1.1372,
      "step": 4644
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.135175883769989,
      "learning_rate": 5.96460132783776e-06,
      "loss": 1.0931,
      "step": 4645
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.14117993414402008,
      "learning_rate": 5.943418198665252e-06,
      "loss": 1.172,
      "step": 4646
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.13064880669116974,
      "learning_rate": 5.922271599985285e-06,
      "loss": 1.1064,
      "step": 4647
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.11799082159996033,
      "learning_rate": 5.901161540010968e-06,
      "loss": 1.1128,
      "step": 4648
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.11777452379465103,
      "learning_rate": 5.8800880269412326e-06,
      "loss": 1.1543,
      "step": 4649
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.17586584389209747,
      "learning_rate": 5.8590510689607415e-06,
      "loss": 1.0077,
      "step": 4650
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1560041457414627,
      "learning_rate": 5.838050674240025e-06,
      "loss": 1.0697,
      "step": 4651
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.143063023686409,
      "learning_rate": 5.817086850935416e-06,
      "loss": 1.0685,
      "step": 4652
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.15187066793441772,
      "learning_rate": 5.796159607189e-06,
      "loss": 1.2016,
      "step": 4653
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13278786838054657,
      "learning_rate": 5.775268951128676e-06,
      "loss": 1.1661,
      "step": 4654
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.11069585382938385,
      "learning_rate": 5.754414890868154e-06,
      "loss": 1.1119,
      "step": 4655
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1253935694694519,
      "learning_rate": 5.733597434506877e-06,
      "loss": 1.102,
      "step": 4656
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.10122222453355789,
      "learning_rate": 5.712816590130132e-06,
      "loss": 1.0576,
      "step": 4657
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13499771058559418,
      "learning_rate": 5.6920723658089534e-06,
      "loss": 1.0799,
      "step": 4658
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.12107738107442856,
      "learning_rate": 5.671364769600163e-06,
      "loss": 1.1137,
      "step": 4659
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1426776796579361,
      "learning_rate": 5.650693809546348e-06,
      "loss": 1.0588,
      "step": 4660
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.16555751860141754,
      "learning_rate": 5.630059493675865e-06,
      "loss": 1.1241,
      "step": 4661
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13491079211235046,
      "learning_rate": 5.609461830002849e-06,
      "loss": 1.1443,
      "step": 4662
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.12806056439876556,
      "learning_rate": 5.5889008265271995e-06,
      "loss": 1.1247,
      "step": 4663
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.10385578870773315,
      "learning_rate": 5.5683764912345615e-06,
      "loss": 1.0046,
      "step": 4664
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.16606464982032776,
      "learning_rate": 5.547888832096382e-06,
      "loss": 1.1011,
      "step": 4665
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13546118140220642,
      "learning_rate": 5.527437857069784e-06,
      "loss": 1.0528,
      "step": 4666
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13527540862560272,
      "learning_rate": 5.507023574097725e-06,
      "loss": 1.0914,
      "step": 4667
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.10587766766548157,
      "learning_rate": 5.486645991108874e-06,
      "loss": 1.1523,
      "step": 4668
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.11526206135749817,
      "learning_rate": 5.4663051160176225e-06,
      "loss": 1.077,
      "step": 4669
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.09295862913131714,
      "learning_rate": 5.446000956724173e-06,
      "loss": 1.1124,
      "step": 4670
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13171526789665222,
      "learning_rate": 5.4257335211143955e-06,
      "loss": 1.0889,
      "step": 4671
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.15201137959957123,
      "learning_rate": 5.4055028170599374e-06,
      "loss": 1.2049,
      "step": 4672
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.11656748503446579,
      "learning_rate": 5.3853088524181914e-06,
      "loss": 1.0951,
      "step": 4673
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.139939084649086,
      "learning_rate": 5.365151635032218e-06,
      "loss": 1.1398,
      "step": 4674
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.11429877579212189,
      "learning_rate": 5.345031172730875e-06,
      "loss": 1.1087,
      "step": 4675
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.19478681683540344,
      "learning_rate": 5.3249474733287345e-06,
      "loss": 1.0417,
      "step": 4676
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.12051262706518173,
      "learning_rate": 5.3049005446260455e-06,
      "loss": 1.1406,
      "step": 4677
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13371168076992035,
      "learning_rate": 5.284890394408826e-06,
      "loss": 1.2234,
      "step": 4678
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1339193433523178,
      "learning_rate": 5.2649170304487576e-06,
      "loss": 1.151,
      "step": 4679
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13793513178825378,
      "learning_rate": 5.2449804605032685e-06,
      "loss": 1.214,
      "step": 4680
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13616138696670532,
      "learning_rate": 5.225080692315532e-06,
      "loss": 1.1265,
      "step": 4681
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.14277154207229614,
      "learning_rate": 5.205217733614353e-06,
      "loss": 1.1157,
      "step": 4682
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.10786115378141403,
      "learning_rate": 5.185391592114286e-06,
      "loss": 1.0947,
      "step": 4683
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.14552143216133118,
      "learning_rate": 5.165602275515591e-06,
      "loss": 1.1501,
      "step": 4684
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.14480701088905334,
      "learning_rate": 5.145849791504187e-06,
      "loss": 1.1099,
      "step": 4685
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13783177733421326,
      "learning_rate": 5.126134147751716e-06,
      "loss": 1.1451,
      "step": 4686
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1290561705827713,
      "learning_rate": 5.106455351915507e-06,
      "loss": 1.0875,
      "step": 4687
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1651766002178192,
      "learning_rate": 5.0868134116385805e-06,
      "loss": 1.0985,
      "step": 4688
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.11932969838380814,
      "learning_rate": 5.067208334549656e-06,
      "loss": 1.1493,
      "step": 4689
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.16001403331756592,
      "learning_rate": 5.047640128263087e-06,
      "loss": 1.1112,
      "step": 4690
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.14310099184513092,
      "learning_rate": 5.028108800378961e-06,
      "loss": 1.154,
      "step": 4691
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1145976111292839,
      "learning_rate": 5.008614358483022e-06,
      "loss": 1.1034,
      "step": 4692
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.16080273687839508,
      "learning_rate": 4.9891568101466666e-06,
      "loss": 1.0768,
      "step": 4693
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.12953291833400726,
      "learning_rate": 4.969736162927019e-06,
      "loss": 1.0976,
      "step": 4694
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.12647269666194916,
      "learning_rate": 4.9503524243667994e-06,
      "loss": 1.0811,
      "step": 4695
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.09254682064056396,
      "learning_rate": 4.931005601994432e-06,
      "loss": 1.0963,
      "step": 4696
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.185682013630867,
      "learning_rate": 4.911695703324038e-06,
      "loss": 1.1521,
      "step": 4697
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13329312205314636,
      "learning_rate": 4.892422735855285e-06,
      "loss": 1.1012,
      "step": 4698
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1041540876030922,
      "learning_rate": 4.873186707073663e-06,
      "loss": 1.1892,
      "step": 4699
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.14535923302173615,
      "learning_rate": 4.853987624450151e-06,
      "loss": 1.0933,
      "step": 4700
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.11723443865776062,
      "learning_rate": 4.834825495441475e-06,
      "loss": 1.1006,
      "step": 4701
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.17701607942581177,
      "learning_rate": 4.815700327490014e-06,
      "loss": 1.1626,
      "step": 4702
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.13782213628292084,
      "learning_rate": 4.796612128023725e-06,
      "loss": 1.1808,
      "step": 4703
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.13317404687404633,
      "learning_rate": 4.777560904456235e-06,
      "loss": 1.0787,
      "step": 4704
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.14426755905151367,
      "learning_rate": 4.758546664186869e-06,
      "loss": 1.0896,
      "step": 4705
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12766233086585999,
      "learning_rate": 4.739569414600498e-06,
      "loss": 1.0846,
      "step": 4706
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2233331948518753,
      "learning_rate": 4.720629163067703e-06,
      "loss": 1.1672,
      "step": 4707
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.17518863081932068,
      "learning_rate": 4.70172591694461e-06,
      "loss": 1.1084,
      "step": 4708
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.13540153205394745,
      "learning_rate": 4.682859683573048e-06,
      "loss": 1.1376,
      "step": 4709
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.14202556014060974,
      "learning_rate": 4.664030470280467e-06,
      "loss": 1.1419,
      "step": 4710
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.13903173804283142,
      "learning_rate": 4.645238284379883e-06,
      "loss": 1.1382,
      "step": 4711
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1449759304523468,
      "learning_rate": 4.6264831331699675e-06,
      "loss": 1.1169,
      "step": 4712
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12448640167713165,
      "learning_rate": 4.60776502393504e-06,
      "loss": 1.1367,
      "step": 4713
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1284894049167633,
      "learning_rate": 4.589083963944951e-06,
      "loss": 1.1013,
      "step": 4714
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.13583390414714813,
      "learning_rate": 4.570439960455242e-06,
      "loss": 1.1392,
      "step": 4715
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1431095004081726,
      "learning_rate": 4.551833020707008e-06,
      "loss": 1.1135,
      "step": 4716
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.14144960045814514,
      "learning_rate": 4.533263151926981e-06,
      "loss": 1.0919,
      "step": 4717
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.14353550970554352,
      "learning_rate": 4.5147303613275016e-06,
      "loss": 1.0473,
      "step": 4718
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.14868617057800293,
      "learning_rate": 4.496234656106457e-06,
      "loss": 1.0807,
      "step": 4719
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12591594457626343,
      "learning_rate": 4.477776043447379e-06,
      "loss": 1.0382,
      "step": 4720
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.14950087666511536,
      "learning_rate": 4.459354530519377e-06,
      "loss": 1.129,
      "step": 4721
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.11909469962120056,
      "learning_rate": 4.440970124477173e-06,
      "loss": 1.0612,
      "step": 4722
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.13619765639305115,
      "learning_rate": 4.422622832461054e-06,
      "loss": 1.1001,
      "step": 4723
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1243291050195694,
      "learning_rate": 4.404312661596877e-06,
      "loss": 1.1236,
      "step": 4724
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12531569600105286,
      "learning_rate": 4.386039618996119e-06,
      "loss": 1.086,
      "step": 4725
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1361588090658188,
      "learning_rate": 4.367803711755814e-06,
      "loss": 1.1111,
      "step": 4726
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1325242817401886,
      "learning_rate": 4.349604946958563e-06,
      "loss": 1.0698,
      "step": 4727
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1524597406387329,
      "learning_rate": 4.331443331672591e-06,
      "loss": 1.146,
      "step": 4728
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.10883516818284988,
      "learning_rate": 4.313318872951633e-06,
      "loss": 1.1565,
      "step": 4729
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12881572544574738,
      "learning_rate": 4.295231577835024e-06,
      "loss": 1.077,
      "step": 4730
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.10448172688484192,
      "learning_rate": 4.277181453347678e-06,
      "loss": 1.1734,
      "step": 4731
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1416001319885254,
      "learning_rate": 4.259168506500022e-06,
      "loss": 1.1063,
      "step": 4732
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.13892999291419983,
      "learning_rate": 4.241192744288091e-06,
      "loss": 1.0286,
      "step": 4733
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1693229079246521,
      "learning_rate": 4.223254173693492e-06,
      "loss": 1.1061,
      "step": 4734
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1631217896938324,
      "learning_rate": 4.205352801683327e-06,
      "loss": 1.1688,
      "step": 4735
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.16502098739147186,
      "learning_rate": 4.187488635210302e-06,
      "loss": 1.1141,
      "step": 4736
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.17338529229164124,
      "learning_rate": 4.1696616812126335e-06,
      "loss": 1.0195,
      "step": 4737
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.14635293185710907,
      "learning_rate": 4.151871946614116e-06,
      "loss": 1.1024,
      "step": 4738
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.11888878792524338,
      "learning_rate": 4.13411943832408e-06,
      "loss": 1.1072,
      "step": 4739
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.16542530059814453,
      "learning_rate": 4.116404163237386e-06,
      "loss": 1.0797,
      "step": 4740
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12796063721179962,
      "learning_rate": 4.098726128234443e-06,
      "loss": 1.1046,
      "step": 4741
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.21040412783622742,
      "learning_rate": 4.081085340181223e-06,
      "loss": 1.2173,
      "step": 4742
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.14348331093788147,
      "learning_rate": 4.06348180592917e-06,
      "loss": 1.1409,
      "step": 4743
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12909826636314392,
      "learning_rate": 4.0459155323153035e-06,
      "loss": 1.1052,
      "step": 4744
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.11330420523881912,
      "learning_rate": 4.028386526162175e-06,
      "loss": 1.197,
      "step": 4745
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12705585360527039,
      "learning_rate": 4.010894794277831e-06,
      "loss": 1.1663,
      "step": 4746
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12770867347717285,
      "learning_rate": 3.993440343455878e-06,
      "loss": 1.1055,
      "step": 4747
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.13546527922153473,
      "learning_rate": 3.976023180475397e-06,
      "loss": 1.0596,
      "step": 4748
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.15695810317993164,
      "learning_rate": 3.95864331210104e-06,
      "loss": 1.064,
      "step": 4749
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1422671377658844,
      "learning_rate": 3.941300745082932e-06,
      "loss": 1.0933,
      "step": 4750
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1319160908460617,
      "learning_rate": 3.923995486156717e-06,
      "loss": 1.0973,
      "step": 4751
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12025977671146393,
      "learning_rate": 3.906727542043598e-06,
      "loss": 1.1744,
      "step": 4752
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.11558841168880463,
      "learning_rate": 3.889496919450208e-06,
      "loss": 1.1236,
      "step": 4753
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.15613515675067902,
      "learning_rate": 3.872303625068729e-06,
      "loss": 1.1322,
      "step": 4754
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.14040173590183258,
      "learning_rate": 3.8551476655768525e-06,
      "loss": 1.1163,
      "step": 4755
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11201713234186172,
      "learning_rate": 3.8380290476377255e-06,
      "loss": 1.1346,
      "step": 4756
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11844584345817566,
      "learning_rate": 3.820947777900063e-06,
      "loss": 1.052,
      "step": 4757
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14234596490859985,
      "learning_rate": 3.803903862998004e-06,
      "loss": 1.1247,
      "step": 4758
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1581065058708191,
      "learning_rate": 3.7868973095512185e-06,
      "loss": 1.2417,
      "step": 4759
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.12132729589939117,
      "learning_rate": 3.7699281241648564e-06,
      "loss": 1.092,
      "step": 4760
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1518484652042389,
      "learning_rate": 3.7529963134295465e-06,
      "loss": 1.1314,
      "step": 4761
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.13281717896461487,
      "learning_rate": 3.7361018839214057e-06,
      "loss": 1.1829,
      "step": 4762
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14026367664337158,
      "learning_rate": 3.7192448422020743e-06,
      "loss": 1.1441,
      "step": 4763
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.10811327397823334,
      "learning_rate": 3.7024251948185816e-06,
      "loss": 1.1617,
      "step": 4764
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14544960856437683,
      "learning_rate": 3.6856429483035028e-06,
      "loss": 1.0789,
      "step": 4765
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11902349442243576,
      "learning_rate": 3.66889810917489e-06,
      "loss": 1.1769,
      "step": 4766
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.19095821678638458,
      "learning_rate": 3.6521906839362185e-06,
      "loss": 1.1654,
      "step": 4767
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1350233405828476,
      "learning_rate": 3.635520679076465e-06,
      "loss": 1.1284,
      "step": 4768
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.15700365602970123,
      "learning_rate": 3.6188881010700726e-06,
      "loss": 1.1413,
      "step": 4769
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.12887093424797058,
      "learning_rate": 3.6022929563769513e-06,
      "loss": 1.1772,
      "step": 4770
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.12514221668243408,
      "learning_rate": 3.585735251442457e-06,
      "loss": 1.1005,
      "step": 4771
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.20471999049186707,
      "learning_rate": 3.5692149926974006e-06,
      "loss": 1.1362,
      "step": 4772
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14188304543495178,
      "learning_rate": 3.552732186558072e-06,
      "loss": 1.1207,
      "step": 4773
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14520040154457092,
      "learning_rate": 3.536286839426195e-06,
      "loss": 1.1926,
      "step": 4774
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14481264352798462,
      "learning_rate": 3.5198789576889603e-06,
      "loss": 1.139,
      "step": 4775
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.15759646892547607,
      "learning_rate": 3.5035085477190143e-06,
      "loss": 1.0859,
      "step": 4776
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.15801073610782623,
      "learning_rate": 3.4871756158744052e-06,
      "loss": 1.1393,
      "step": 4777
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.16501332819461823,
      "learning_rate": 3.470880168498669e-06,
      "loss": 1.0457,
      "step": 4778
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14966781437397003,
      "learning_rate": 3.454622211920766e-06,
      "loss": 1.1928,
      "step": 4779
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11361715197563171,
      "learning_rate": 3.4384017524551116e-06,
      "loss": 1.1046,
      "step": 4780
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1551242172718048,
      "learning_rate": 3.422218796401544e-06,
      "loss": 1.1421,
      "step": 4781
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.12656645476818085,
      "learning_rate": 3.4060733500453247e-06,
      "loss": 1.0922,
      "step": 4782
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1435282677412033,
      "learning_rate": 3.3899654196571597e-06,
      "loss": 1.0491,
      "step": 4783
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.16075146198272705,
      "learning_rate": 3.3738950114932e-06,
      "loss": 0.9896,
      "step": 4784
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11541453003883362,
      "learning_rate": 3.3578621317949756e-06,
      "loss": 1.2147,
      "step": 4785
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1326826810836792,
      "learning_rate": 3.3418667867895047e-06,
      "loss": 1.1175,
      "step": 4786
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.19838736951351166,
      "learning_rate": 3.3259089826891854e-06,
      "loss": 1.075,
      "step": 4787
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.13734717667102814,
      "learning_rate": 3.309988725691837e-06,
      "loss": 1.1152,
      "step": 4788
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11212459951639175,
      "learning_rate": 3.294106021980714e-06,
      "loss": 1.1183,
      "step": 4789
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.18137459456920624,
      "learning_rate": 3.2782608777244705e-06,
      "loss": 0.9935,
      "step": 4790
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11466678231954575,
      "learning_rate": 3.2624532990771505e-06,
      "loss": 1.0591,
      "step": 4791
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.12196754664182663,
      "learning_rate": 3.2466832921782985e-06,
      "loss": 1.1005,
      "step": 4792
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11746000498533249,
      "learning_rate": 3.2309508631527486e-06,
      "loss": 1.1967,
      "step": 4793
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.12605242431163788,
      "learning_rate": 3.215256018110824e-06,
      "loss": 1.1279,
      "step": 4794
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.12071791291236877,
      "learning_rate": 3.1995987631482148e-06,
      "loss": 1.211,
      "step": 4795
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14646035432815552,
      "learning_rate": 3.1839791043460023e-06,
      "loss": 1.11,
      "step": 4796
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.13723509013652802,
      "learning_rate": 3.168397047770699e-06,
      "loss": 1.0507,
      "step": 4797
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.13553927838802338,
      "learning_rate": 3.1528525994741877e-06,
      "loss": 1.1222,
      "step": 4798
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11683168262243271,
      "learning_rate": 3.13734576549376e-06,
      "loss": 1.0982,
      "step": 4799
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.11230519413948059,
      "learning_rate": 3.1218765518520985e-06,
      "loss": 1.0818,
      "step": 4800
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1235259622335434,
      "learning_rate": 3.1064449645572537e-06,
      "loss": 1.0745,
      "step": 4801
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.10538920015096664,
      "learning_rate": 3.0910510096026745e-06,
      "loss": 1.0597,
      "step": 4802
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.18483448028564453,
      "learning_rate": 3.0756946929672014e-06,
      "loss": 1.067,
      "step": 4803
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14214687049388885,
      "learning_rate": 3.0603760206150524e-06,
      "loss": 1.0796,
      "step": 4804
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.15218165516853333,
      "learning_rate": 3.045094998495834e-06,
      "loss": 1.1681,
      "step": 4805
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.12442541122436523,
      "learning_rate": 3.029851632544489e-06,
      "loss": 1.1976,
      "step": 4806
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.14230991899967194,
      "learning_rate": 3.0146459286813923e-06,
      "loss": 1.1608,
      "step": 4807
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.10619164258241653,
      "learning_rate": 2.9994778928122636e-06,
      "loss": 1.2031,
      "step": 4808
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1455889195203781,
      "learning_rate": 2.9843475308281575e-06,
      "loss": 1.1386,
      "step": 4809
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.13575752079486847,
      "learning_rate": 2.9692548486055847e-06,
      "loss": 1.2258,
      "step": 4810
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.15925909578800201,
      "learning_rate": 2.954199852006334e-06,
      "loss": 1.1195,
      "step": 4811
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.11832872778177261,
      "learning_rate": 2.9391825468775945e-06,
      "loss": 1.1316,
      "step": 4812
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.11214614659547806,
      "learning_rate": 2.9242029390519456e-06,
      "loss": 1.141,
      "step": 4813
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1280299872159958,
      "learning_rate": 2.909261034347255e-06,
      "loss": 1.1097,
      "step": 4814
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1306041181087494,
      "learning_rate": 2.8943568385667918e-06,
      "loss": 1.0357,
      "step": 4815
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1456451714038849,
      "learning_rate": 2.8794903574992037e-06,
      "loss": 0.994,
      "step": 4816
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.14572738111019135,
      "learning_rate": 2.8646615969184276e-06,
      "loss": 1.1316,
      "step": 4817
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.14161597192287445,
      "learning_rate": 2.8498705625838117e-06,
      "loss": 1.1345,
      "step": 4818
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.12698370218276978,
      "learning_rate": 2.8351172602399946e-06,
      "loss": 1.2075,
      "step": 4819
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.14146335422992706,
      "learning_rate": 2.820401695616992e-06,
      "loss": 1.197,
      "step": 4820
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.16925986111164093,
      "learning_rate": 2.8057238744301996e-06,
      "loss": 1.0425,
      "step": 4821
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.12454888224601746,
      "learning_rate": 2.7910838023802676e-06,
      "loss": 1.0783,
      "step": 4822
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.16842064261436462,
      "learning_rate": 2.7764814851532482e-06,
      "loss": 1.0917,
      "step": 4823
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.12724734842777252,
      "learning_rate": 2.761916928420527e-06,
      "loss": 1.0796,
      "step": 4824
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.13600929081439972,
      "learning_rate": 2.74739013783879e-06,
      "loss": 1.1514,
      "step": 4825
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1350419521331787,
      "learning_rate": 2.7329011190500796e-06,
      "loss": 1.1255,
      "step": 4826
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.12883150577545166,
      "learning_rate": 2.7184498776817612e-06,
      "loss": 1.0884,
      "step": 4827
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.16649560630321503,
      "learning_rate": 2.704036419346534e-06,
      "loss": 1.0565,
      "step": 4828
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.13686098158359528,
      "learning_rate": 2.689660749642442e-06,
      "loss": 1.1783,
      "step": 4829
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.12019894272089005,
      "learning_rate": 2.675322874152786e-06,
      "loss": 1.137,
      "step": 4830
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.16154630482196808,
      "learning_rate": 2.6610227984462554e-06,
      "loss": 1.1355,
      "step": 4831
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.13179606199264526,
      "learning_rate": 2.6467605280768413e-06,
      "loss": 1.1094,
      "step": 4832
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.0989982858300209,
      "learning_rate": 2.632536068583824e-06,
      "loss": 1.0731,
      "step": 4833
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.126750648021698,
      "learning_rate": 2.6183494254918505e-06,
      "loss": 1.1171,
      "step": 4834
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1477127969264984,
      "learning_rate": 2.6042006043108247e-06,
      "loss": 1.2215,
      "step": 4835
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.14067818224430084,
      "learning_rate": 2.590089610535984e-06,
      "loss": 1.1916,
      "step": 4836
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.22255636751651764,
      "learning_rate": 2.5760164496479e-06,
      "loss": 1.1117,
      "step": 4837
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.12695303559303284,
      "learning_rate": 2.5619811271123894e-06,
      "loss": 1.0939,
      "step": 4838
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1740741729736328,
      "learning_rate": 2.5479836483806587e-06,
      "loss": 1.1047,
      "step": 4839
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.09920041263103485,
      "learning_rate": 2.534024018889114e-06,
      "loss": 1.1093,
      "step": 4840
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1257711946964264,
      "learning_rate": 2.5201022440595522e-06,
      "loss": 1.1127,
      "step": 4841
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.13650232553482056,
      "learning_rate": 2.506218329299026e-06,
      "loss": 1.1666,
      "step": 4842
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.15982040762901306,
      "learning_rate": 2.492372279999866e-06,
      "loss": 1.0747,
      "step": 4843
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.11588667333126068,
      "learning_rate": 2.4785641015397374e-06,
      "loss": 1.1085,
      "step": 4844
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.16456179320812225,
      "learning_rate": 2.464793799281573e-06,
      "loss": 1.0747,
      "step": 4845
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.17404890060424805,
      "learning_rate": 2.4510613785735938e-06,
      "loss": 1.1482,
      "step": 4846
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1565217524766922,
      "learning_rate": 2.4373668447493224e-06,
      "loss": 1.1424,
      "step": 4847
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.125652015209198,
      "learning_rate": 2.4237102031275607e-06,
      "loss": 1.1377,
      "step": 4848
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.14873909950256348,
      "learning_rate": 2.4100914590123756e-06,
      "loss": 1.1006,
      "step": 4849
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1197202205657959,
      "learning_rate": 2.3965106176931374e-06,
      "loss": 1.274,
      "step": 4850
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.14618109166622162,
      "learning_rate": 2.382967684444493e-06,
      "loss": 1.1259,
      "step": 4851
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.11620212346315384,
      "learning_rate": 2.3694626645263675e-06,
      "loss": 1.1066,
      "step": 4852
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.13322843611240387,
      "learning_rate": 2.3559955631839434e-06,
      "loss": 1.0871,
      "step": 4853
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1501568704843521,
      "learning_rate": 2.3425663856476932e-06,
      "loss": 1.074,
      "step": 4854
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.12066781520843506,
      "learning_rate": 2.3291751371333437e-06,
      "loss": 1.0915,
      "step": 4855
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.14910274744033813,
      "learning_rate": 2.3158218228419127e-06,
      "loss": 1.0661,
      "step": 4856
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.12110350281000137,
      "learning_rate": 2.3025064479596627e-06,
      "loss": 1.0998,
      "step": 4857
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.11612378060817719,
      "learning_rate": 2.2892290176581676e-06,
      "loss": 1.134,
      "step": 4858
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.14989136159420013,
      "learning_rate": 2.2759895370941807e-06,
      "loss": 1.1675,
      "step": 4859
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1220552995800972,
      "learning_rate": 2.2627880114097776e-06,
      "loss": 1.0851,
      "step": 4860
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1535583734512329,
      "learning_rate": 2.249624445732301e-06,
      "loss": 1.0634,
      "step": 4861
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.17036940157413483,
      "learning_rate": 2.2364988451742953e-06,
      "loss": 1.1152,
      "step": 4862
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1345413327217102,
      "learning_rate": 2.223411214833637e-06,
      "loss": 1.1116,
      "step": 4863
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.15457816421985626,
      "learning_rate": 2.210361559793361e-06,
      "loss": 1.0496,
      "step": 4864
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1472393274307251,
      "learning_rate": 2.1973498851218242e-06,
      "loss": 1.0737,
      "step": 4865
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.16097760200500488,
      "learning_rate": 2.1843761958726285e-06,
      "loss": 0.973,
      "step": 4866
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.15206554532051086,
      "learning_rate": 2.1714404970845647e-06,
      "loss": 1.2155,
      "step": 4867
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1420641392469406,
      "learning_rate": 2.158542793781759e-06,
      "loss": 1.1195,
      "step": 4868
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2195146679878235,
      "learning_rate": 2.1456830909734936e-06,
      "loss": 1.14,
      "step": 4869
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1315392702817917,
      "learning_rate": 2.1328613936543397e-06,
      "loss": 1.0793,
      "step": 4870
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1737673580646515,
      "learning_rate": 2.1200777068041134e-06,
      "loss": 1.1089,
      "step": 4871
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.11330298334360123,
      "learning_rate": 2.10733203538781e-06,
      "loss": 1.045,
      "step": 4872
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.12656517326831818,
      "learning_rate": 2.0946243843557367e-06,
      "loss": 1.1975,
      "step": 4873
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1265537589788437,
      "learning_rate": 2.081954758643401e-06,
      "loss": 1.0932,
      "step": 4874
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1523357331752777,
      "learning_rate": 2.0693231631715103e-06,
      "loss": 1.1634,
      "step": 4875
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1692282259464264,
      "learning_rate": 2.0567296028460637e-06,
      "loss": 1.1353,
      "step": 4876
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.13702724874019623,
      "learning_rate": 2.0441740825582257e-06,
      "loss": 1.175,
      "step": 4877
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.10840611159801483,
      "learning_rate": 2.03165660718444e-06,
      "loss": 1.0623,
      "step": 4878
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.12313251197338104,
      "learning_rate": 2.019177181586329e-06,
      "loss": 1.0849,
      "step": 4879
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.14894047379493713,
      "learning_rate": 2.0067358106107716e-06,
      "loss": 1.1499,
      "step": 4880
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.10263068974018097,
      "learning_rate": 1.994332499089846e-06,
      "loss": 1.141,
      "step": 4881
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.11073864251375198,
      "learning_rate": 1.9819672518408663e-06,
      "loss": 1.0646,
      "step": 4882
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.16947859525680542,
      "learning_rate": 1.9696400736663457e-06,
      "loss": 1.2308,
      "step": 4883
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.11136303097009659,
      "learning_rate": 1.9573509693540102e-06,
      "loss": 1.0798,
      "step": 4884
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1377623826265335,
      "learning_rate": 1.945099943676809e-06,
      "loss": 1.1739,
      "step": 4885
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.14100700616836548,
      "learning_rate": 1.9328870013929135e-06,
      "loss": 1.1585,
      "step": 4886
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.15513524413108826,
      "learning_rate": 1.9207121472456847e-06,
      "loss": 1.1779,
      "step": 4887
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1244889572262764,
      "learning_rate": 1.9085753859636734e-06,
      "loss": 1.0307,
      "step": 4888
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.12940151989459991,
      "learning_rate": 1.8964767222606871e-06,
      "loss": 1.0511,
      "step": 4889
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1724206805229187,
      "learning_rate": 1.8844161608356781e-06,
      "loss": 1.1079,
      "step": 4890
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.12597207725048065,
      "learning_rate": 1.872393706372866e-06,
      "loss": 1.1589,
      "step": 4891
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.14497388899326324,
      "learning_rate": 1.8604093635416153e-06,
      "loss": 1.149,
      "step": 4892
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.134937584400177,
      "learning_rate": 1.8484631369964922e-06,
      "loss": 1.0689,
      "step": 4893
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.14417286217212677,
      "learning_rate": 1.836555031377285e-06,
      "loss": 1.1262,
      "step": 4894
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.13951830565929413,
      "learning_rate": 1.824685051308983e-06,
      "loss": 1.0568,
      "step": 4895
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.11766871064901352,
      "learning_rate": 1.8128532014017096e-06,
      "loss": 1.0661,
      "step": 4896
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1189696192741394,
      "learning_rate": 1.8010594862508668e-06,
      "loss": 1.1232,
      "step": 4897
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1658710539340973,
      "learning_rate": 1.789303910436968e-06,
      "loss": 1.1929,
      "step": 4898
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.12337343394756317,
      "learning_rate": 1.777586478525739e-06,
      "loss": 1.069,
      "step": 4899
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.10449209064245224,
      "learning_rate": 1.7659071950681172e-06,
      "loss": 1.0773,
      "step": 4900
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.13650374114513397,
      "learning_rate": 1.754266064600174e-06,
      "loss": 1.1371,
      "step": 4901
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.12660443782806396,
      "learning_rate": 1.7426630916432041e-06,
      "loss": 1.0597,
      "step": 4902
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1136551946401596,
      "learning_rate": 1.7310982807036912e-06,
      "loss": 1.068,
      "step": 4903
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1360306590795517,
      "learning_rate": 1.7195716362732427e-06,
      "loss": 1.0796,
      "step": 4904
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1353238970041275,
      "learning_rate": 1.7080831628286887e-06,
      "loss": 1.1107,
      "step": 4905
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1169363483786583,
      "learning_rate": 1.696632864832015e-06,
      "loss": 1.1074,
      "step": 4906
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1420018970966339,
      "learning_rate": 1.685220746730387e-06,
      "loss": 1.0985,
      "step": 4907
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.14447839558124542,
      "learning_rate": 1.6738468129561369e-06,
      "loss": 1.1325,
      "step": 4908
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.15822412073612213,
      "learning_rate": 1.6625110679267641e-06,
      "loss": 1.1126,
      "step": 4909
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1203211322426796,
      "learning_rate": 1.6512135160449583e-06,
      "loss": 1.1542,
      "step": 4910
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.15209484100341797,
      "learning_rate": 1.6399541616985648e-06,
      "loss": 1.1186,
      "step": 4911
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14795251190662384,
      "learning_rate": 1.6287330092605524e-06,
      "loss": 1.1637,
      "step": 4912
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14722613990306854,
      "learning_rate": 1.6175500630891127e-06,
      "loss": 1.0835,
      "step": 4913
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.166097030043602,
      "learning_rate": 1.6064053275275714e-06,
      "loss": 1.0508,
      "step": 4914
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12818530201911926,
      "learning_rate": 1.5952988069044106e-06,
      "loss": 1.078,
      "step": 4915
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.11715701967477798,
      "learning_rate": 1.5842305055332795e-06,
      "loss": 1.1113,
      "step": 4916
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.110464908182621,
      "learning_rate": 1.573200427712973e-06,
      "loss": 1.1737,
      "step": 4917
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.11091790348291397,
      "learning_rate": 1.562208577727442e-06,
      "loss": 1.0864,
      "step": 4918
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.18288132548332214,
      "learning_rate": 1.5512549598458048e-06,
      "loss": 1.1601,
      "step": 4919
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.15630561113357544,
      "learning_rate": 1.540339578322314e-06,
      "loss": 1.1454,
      "step": 4920
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14643220603466034,
      "learning_rate": 1.5294624373963895e-06,
      "loss": 1.1239,
      "step": 4921
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14837227761745453,
      "learning_rate": 1.5186235412925742e-06,
      "loss": 1.1334,
      "step": 4922
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1300095021724701,
      "learning_rate": 1.5078228942205674e-06,
      "loss": 1.0394,
      "step": 4923
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1361258625984192,
      "learning_rate": 1.497060500375236e-06,
      "loss": 1.1209,
      "step": 4924
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14800000190734863,
      "learning_rate": 1.4863363639365357e-06,
      "loss": 1.0675,
      "step": 4925
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.16090898215770721,
      "learning_rate": 1.4756504890696466e-06,
      "loss": 1.1494,
      "step": 4926
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1313980668783188,
      "learning_rate": 1.4650028799247928e-06,
      "loss": 1.1392,
      "step": 4927
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12921400368213654,
      "learning_rate": 1.454393540637411e-06,
      "loss": 1.2106,
      "step": 4928
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.10059855133295059,
      "learning_rate": 1.4438224753280382e-06,
      "loss": 1.0589,
      "step": 4929
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14263777434825897,
      "learning_rate": 1.4332896881023461e-06,
      "loss": 1.0763,
      "step": 4930
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12562771141529083,
      "learning_rate": 1.422795183051151e-06,
      "loss": 1.1371,
      "step": 4931
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.17867805063724518,
      "learning_rate": 1.4123389642504148e-06,
      "loss": 1.187,
      "step": 4932
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.13355132937431335,
      "learning_rate": 1.401921035761189e-06,
      "loss": 1.0892,
      "step": 4933
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.15874680876731873,
      "learning_rate": 1.3915414016296924e-06,
      "loss": 1.0245,
      "step": 4934
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.21350693702697754,
      "learning_rate": 1.381200065887256e-06,
      "loss": 1.1431,
      "step": 4935
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.13366438448429108,
      "learning_rate": 1.3708970325503223e-06,
      "loss": 1.1146,
      "step": 4936
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12210503220558167,
      "learning_rate": 1.3606323056204795e-06,
      "loss": 1.1488,
      "step": 4937
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.15479691326618195,
      "learning_rate": 1.3504058890844273e-06,
      "loss": 1.0536,
      "step": 4938
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.10433192551136017,
      "learning_rate": 1.3402177869139886e-06,
      "loss": 1.044,
      "step": 4939
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.11626780778169632,
      "learning_rate": 1.3300680030661095e-06,
      "loss": 1.1835,
      "step": 4940
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12838445603847504,
      "learning_rate": 1.3199565414828364e-06,
      "loss": 1.1512,
      "step": 4941
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1496628075838089,
      "learning_rate": 1.309883406091361e-06,
      "loss": 1.1238,
      "step": 4942
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.15561267733573914,
      "learning_rate": 1.2998486008039545e-06,
      "loss": 1.0607,
      "step": 4943
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.11093346029520035,
      "learning_rate": 1.2898521295180322e-06,
      "loss": 1.1187,
      "step": 4944
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.13317137956619263,
      "learning_rate": 1.2798939961161215e-06,
      "loss": 1.2068,
      "step": 4945
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14453087747097015,
      "learning_rate": 1.269974204465818e-06,
      "loss": 1.1737,
      "step": 4946
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.15410009026527405,
      "learning_rate": 1.260092758419862e-06,
      "loss": 1.1621,
      "step": 4947
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.13580329716205597,
      "learning_rate": 1.2502496618161164e-06,
      "loss": 1.1026,
      "step": 4948
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12090010195970535,
      "learning_rate": 1.2404449184774902e-06,
      "loss": 1.1217,
      "step": 4949
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1356223076581955,
      "learning_rate": 1.2306785322120595e-06,
      "loss": 1.1316,
      "step": 4950
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.16575613617897034,
      "learning_rate": 1.220950506812968e-06,
      "loss": 1.097,
      "step": 4951
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12525203824043274,
      "learning_rate": 1.2112608460584707e-06,
      "loss": 1.087,
      "step": 4952
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1214626133441925,
      "learning_rate": 1.2016095537119243e-06,
      "loss": 1.1074,
      "step": 4953
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14401765167713165,
      "learning_rate": 1.1919966335217636e-06,
      "loss": 1.0757,
      "step": 4954
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.0995514914393425,
      "learning_rate": 1.1824220892215464e-06,
      "loss": 1.1126,
      "step": 4955
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.11953631788492203,
      "learning_rate": 1.1728859245299205e-06,
      "loss": 1.1087,
      "step": 4956
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1268925964832306,
      "learning_rate": 1.1633881431506122e-06,
      "loss": 1.132,
      "step": 4957
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14771179854869843,
      "learning_rate": 1.1539287487724593e-06,
      "loss": 1.1296,
      "step": 4958
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.2848549783229828,
      "learning_rate": 1.1445077450693786e-06,
      "loss": 1.1174,
      "step": 4959
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12932825088500977,
      "learning_rate": 1.1351251357003655e-06,
      "loss": 1.167,
      "step": 4960
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.14389464259147644,
      "learning_rate": 1.1257809243095386e-06,
      "loss": 1.0993,
      "step": 4961
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.11870590597391129,
      "learning_rate": 1.1164751145260722e-06,
      "loss": 1.0837,
      "step": 4962
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.12540219724178314,
      "learning_rate": 1.1072077099642418e-06,
      "loss": 1.0166,
      "step": 4963
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.11562046408653259,
      "learning_rate": 1.097978714223391e-06,
      "loss": 1.0903,
      "step": 4964
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.17831215262413025,
      "learning_rate": 1.0887881308879633e-06,
      "loss": 1.1397,
      "step": 4965
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14712804555892944,
      "learning_rate": 1.07963596352747e-06,
      "loss": 1.134,
      "step": 4966
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.13710719347000122,
      "learning_rate": 1.0705222156965012e-06,
      "loss": 1.1689,
      "step": 4967
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1609877198934555,
      "learning_rate": 1.0614468909347474e-06,
      "loss": 1.0869,
      "step": 4968
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14530976116657257,
      "learning_rate": 1.0524099927669563e-06,
      "loss": 1.1809,
      "step": 4969
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.12129450589418411,
      "learning_rate": 1.043411524702942e-06,
      "loss": 1.1417,
      "step": 4970
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.16022862493991852,
      "learning_rate": 1.03445149023762e-06,
      "loss": 1.0922,
      "step": 4971
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.11986338347196579,
      "learning_rate": 1.0255298928509627e-06,
      "loss": 1.1736,
      "step": 4972
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.15767605602741241,
      "learning_rate": 1.016646736007998e-06,
      "loss": 1.0852,
      "step": 4973
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.10621188580989838,
      "learning_rate": 1.0078020231588768e-06,
      "loss": 1.1308,
      "step": 4974
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18066060543060303,
      "learning_rate": 9.989957577387521e-07,
      "loss": 1.0633,
      "step": 4975
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14141497015953064,
      "learning_rate": 9.902279431678873e-07,
      "loss": 1.1255,
      "step": 4976
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.13976971805095673,
      "learning_rate": 9.814985828516033e-07,
      "loss": 1.1203,
      "step": 4977
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.11208932846784592,
      "learning_rate": 9.728076801802655e-07,
      "loss": 1.0937,
      "step": 4978
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.12932173907756805,
      "learning_rate": 9.641552385293518e-07,
      "loss": 1.069,
      "step": 4979
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.15481196343898773,
      "learning_rate": 9.555412612593518e-07,
      "loss": 1.0556,
      "step": 4980
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.12664298713207245,
      "learning_rate": 9.469657517158226e-07,
      "loss": 1.0318,
      "step": 4981
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.09144528210163116,
      "learning_rate": 9.384287132294223e-07,
      "loss": 1.1076,
      "step": 4982
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.10765048861503601,
      "learning_rate": 9.299301491158207e-07,
      "loss": 1.1552,
      "step": 4983
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1573924571275711,
      "learning_rate": 9.214700626757666e-07,
      "loss": 1.2124,
      "step": 4984
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.15100307762622833,
      "learning_rate": 9.130484571950537e-07,
      "loss": 1.1515,
      "step": 4985
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.12261990457773209,
      "learning_rate": 9.046653359445323e-07,
      "loss": 1.1658,
      "step": 4986
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.13252341747283936,
      "learning_rate": 8.963207021801423e-07,
      "loss": 1.1071,
      "step": 4987
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.10131005197763443,
      "learning_rate": 8.880145591428024e-07,
      "loss": 1.0783,
      "step": 4988
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.09961023181676865,
      "learning_rate": 8.797469100585431e-07,
      "loss": 1.1607,
      "step": 4989
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.12545712292194366,
      "learning_rate": 8.715177581384182e-07,
      "loss": 1.0637,
      "step": 4990
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.15748906135559082,
      "learning_rate": 8.633271065785486e-07,
      "loss": 1.0888,
      "step": 4991
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14925558865070343,
      "learning_rate": 8.551749585600677e-07,
      "loss": 1.0747,
      "step": 4992
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.12721112370491028,
      "learning_rate": 8.470613172491981e-07,
      "loss": 1.1477,
      "step": 4993
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.149870827794075,
      "learning_rate": 8.389861857971748e-07,
      "loss": 1.099,
      "step": 4994
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1465064287185669,
      "learning_rate": 8.309495673402779e-07,
      "loss": 1.182,
      "step": 4995
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14628492295742035,
      "learning_rate": 8.229514649998437e-07,
      "loss": 1.1915,
      "step": 4996
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14620113372802734,
      "learning_rate": 8.149918818822433e-07,
      "loss": 1.2084,
      "step": 4997
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.15156982839107513,
      "learning_rate": 8.070708210788924e-07,
      "loss": 1.1272,
      "step": 4998
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1706887185573578,
      "learning_rate": 7.991882856662302e-07,
      "loss": 1.1155,
      "step": 4999
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14042803645133972,
      "learning_rate": 7.913442787057523e-07,
      "loss": 1.0965,
      "step": 5000
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14614254236221313,
      "learning_rate": 7.835388032439661e-07,
      "loss": 1.0437,
      "step": 5001
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1262795627117157,
      "learning_rate": 7.757718623124466e-07,
      "loss": 1.0662,
      "step": 5002
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14818024635314941,
      "learning_rate": 7.680434589277697e-07,
      "loss": 1.0647,
      "step": 5003
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1495153307914734,
      "learning_rate": 7.603535960915675e-07,
      "loss": 1.0864,
      "step": 5004
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.09188217669725418,
      "learning_rate": 7.527022767904956e-07,
      "loss": 1.0526,
      "step": 5005
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.11428637057542801,
      "learning_rate": 7.450895039962214e-07,
      "loss": 1.1459,
      "step": 5006
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.16746792197227478,
      "learning_rate": 7.375152806654683e-07,
      "loss": 1.1146,
      "step": 5007
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.10246127098798752,
      "learning_rate": 7.299796097399947e-07,
      "loss": 1.0759,
      "step": 5008
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.13786228001117706,
      "learning_rate": 7.224824941465369e-07,
      "loss": 1.1404,
      "step": 5009
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.13943424820899963,
      "learning_rate": 7.150239367969102e-07,
      "loss": 1.1565,
      "step": 5010
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1219056099653244,
      "learning_rate": 7.076039405879309e-07,
      "loss": 1.206,
      "step": 5011
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14585334062576294,
      "learning_rate": 7.002225084014269e-07,
      "loss": 1.1144,
      "step": 5012
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.14585334062576294,
      "learning_rate": 7.002225084014269e-07,
      "loss": 1.1615,
      "step": 5013
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.13533379137516022,
      "learning_rate": 6.928796431042717e-07,
      "loss": 1.1115,
      "step": 5014
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1648525893688202,
      "learning_rate": 6.855753475483506e-07,
      "loss": 1.0398,
      "step": 5015
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.14393064379692078,
      "learning_rate": 6.783096245705611e-07,
      "loss": 1.1965,
      "step": 5016
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.11794988811016083,
      "learning_rate": 6.710824769928348e-07,
      "loss": 1.103,
      "step": 5017
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.16352829337120056,
      "learning_rate": 6.63893907622104e-07,
      "loss": 1.1221,
      "step": 5018
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.12162108719348907,
      "learning_rate": 6.567439192503245e-07,
      "loss": 1.119,
      "step": 5019
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.12206538766622543,
      "learning_rate": 6.496325146544746e-07,
      "loss": 1.1681,
      "step": 5020
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.14267969131469727,
      "learning_rate": 6.425596965965452e-07,
      "loss": 1.0861,
      "step": 5021
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.11071905493736267,
      "learning_rate": 6.35525467823539e-07,
      "loss": 1.0984,
      "step": 5022
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13089194893836975,
      "learning_rate": 6.285298310674703e-07,
      "loss": 1.1037,
      "step": 5023
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.14431218802928925,
      "learning_rate": 6.215727890453438e-07,
      "loss": 1.0934,
      "step": 5024
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.14242678880691528,
      "learning_rate": 6.14654344459209e-07,
      "loss": 1.1283,
      "step": 5025
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.12703518569469452,
      "learning_rate": 6.077744999961165e-07,
      "loss": 1.118,
      "step": 5026
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1386290341615677,
      "learning_rate": 6.009332583281069e-07,
      "loss": 1.1193,
      "step": 5027
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13291719555854797,
      "learning_rate": 5.941306221122545e-07,
      "loss": 1.1244,
      "step": 5028
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13277825713157654,
      "learning_rate": 5.873665939906015e-07,
      "loss": 1.2037,
      "step": 5029
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.12441828846931458,
      "learning_rate": 5.806411765902353e-07,
      "loss": 1.0913,
      "step": 5030
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1443171352148056,
      "learning_rate": 5.739543725232332e-07,
      "loss": 1.1927,
      "step": 5031
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.17369720339775085,
      "learning_rate": 5.673061843866623e-07,
      "loss": 1.1348,
      "step": 5032
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13111120462417603,
      "learning_rate": 5.606966147626124e-07,
      "loss": 1.0861,
      "step": 5033
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1564483791589737,
      "learning_rate": 5.541256662181526e-07,
      "loss": 1.1885,
      "step": 5034
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13232482969760895,
      "learning_rate": 5.475933413053636e-07,
      "loss": 1.1081,
      "step": 5035
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.12528836727142334,
      "learning_rate": 5.410996425613379e-07,
      "loss": 1.1152,
      "step": 5036
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.12708434462547302,
      "learning_rate": 5.346445725081473e-07,
      "loss": 1.1385,
      "step": 5037
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13817031681537628,
      "learning_rate": 5.282281336528638e-07,
      "loss": 1.1472,
      "step": 5038
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.09404988586902618,
      "learning_rate": 5.218503284875609e-07,
      "loss": 1.1204,
      "step": 5039
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13360686600208282,
      "learning_rate": 5.155111594893014e-07,
      "loss": 1.1376,
      "step": 5040
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13145127892494202,
      "learning_rate": 5.092106291201604e-07,
      "loss": 1.1792,
      "step": 5041
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.11990705132484436,
      "learning_rate": 5.029487398271692e-07,
      "loss": 1.0856,
      "step": 5042
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13673651218414307,
      "learning_rate": 4.967254940423715e-07,
      "loss": 1.1023,
      "step": 5043
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13377781212329865,
      "learning_rate": 4.905408941828338e-07,
      "loss": 1.1236,
      "step": 5044
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1276330202817917,
      "learning_rate": 4.843949426505567e-07,
      "loss": 1.0745,
      "step": 5045
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.09419012814760208,
      "learning_rate": 4.782876418325755e-07,
      "loss": 1.0512,
      "step": 5046
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1675354242324829,
      "learning_rate": 4.7221899410087033e-07,
      "loss": 1.1645,
      "step": 5047
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13166339695453644,
      "learning_rate": 4.6618900181245595e-07,
      "loss": 1.1507,
      "step": 5048
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.12378307431936264,
      "learning_rate": 4.6019766730930336e-07,
      "loss": 1.0365,
      "step": 5049
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.11966212093830109,
      "learning_rate": 4.5424499291838454e-07,
      "loss": 1.1225,
      "step": 5050
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1220046728849411,
      "learning_rate": 4.4833098095165004e-07,
      "loss": 1.1554,
      "step": 5051
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.10551996529102325,
      "learning_rate": 4.4245563370601815e-07,
      "loss": 1.0487,
      "step": 5052
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1294924020767212,
      "learning_rate": 4.366189534634191e-07,
      "loss": 1.0978,
      "step": 5053
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.14619861543178558,
      "learning_rate": 4.308209424907506e-07,
      "loss": 1.1453,
      "step": 5054
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.14435534179210663,
      "learning_rate": 4.250616030399002e-07,
      "loss": 1.224,
      "step": 5055
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.15203510224819183,
      "learning_rate": 4.1934093734771195e-07,
      "loss": 1.1015,
      "step": 5056
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1504582017660141,
      "learning_rate": 4.1365894763604197e-07,
      "loss": 1.2512,
      "step": 5057
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13070113956928253,
      "learning_rate": 4.080156361117027e-07,
      "loss": 1.1319,
      "step": 5058
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13383358716964722,
      "learning_rate": 4.024110049664853e-07,
      "loss": 1.1081,
      "step": 5059
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13002096116542816,
      "learning_rate": 3.9684505637718194e-07,
      "loss": 1.135,
      "step": 5060
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.16403760015964508,
      "learning_rate": 3.913177925055189e-07,
      "loss": 1.0946,
      "step": 5061
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.16107772290706635,
      "learning_rate": 3.858292154982457e-07,
      "loss": 1.1766,
      "step": 5062
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.12790405750274658,
      "learning_rate": 3.8037932748704597e-07,
      "loss": 1.1589,
      "step": 5063
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.11785704642534256,
      "learning_rate": 3.7496813058859326e-07,
      "loss": 1.0948,
      "step": 5064
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.11422199010848999,
      "learning_rate": 3.6959562690455087e-07,
      "loss": 1.0788,
      "step": 5065
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.14197655022144318,
      "learning_rate": 3.642618185215163e-07,
      "loss": 1.074,
      "step": 5066
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.13659080862998962,
      "learning_rate": 3.5896670751109916e-07,
      "loss": 1.106,
      "step": 5067
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.20332708954811096,
      "learning_rate": 3.537102959298322e-07,
      "loss": 1.1498,
      "step": 5068
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.16408969461917877,
      "learning_rate": 3.484925858192822e-07,
      "loss": 1.1607,
      "step": 5069
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.16305966675281525,
      "learning_rate": 3.4331357920591724e-07,
      "loss": 1.103,
      "step": 5070
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.10167058557271957,
      "learning_rate": 3.38173278101217e-07,
      "loss": 1.1658,
      "step": 5071
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12583011388778687,
      "learning_rate": 3.3307168450160685e-07,
      "loss": 1.1689,
      "step": 5072
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.15563678741455078,
      "learning_rate": 3.280088003885018e-07,
      "loss": 1.1529,
      "step": 5073
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12005467712879181,
      "learning_rate": 3.2298462772825113e-07,
      "loss": 1.2313,
      "step": 5074
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.16251133382320404,
      "learning_rate": 3.17999168472205e-07,
      "loss": 1.1546,
      "step": 5075
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12847298383712769,
      "learning_rate": 3.130524245566369e-07,
      "loss": 1.1111,
      "step": 5076
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.11485613137483597,
      "learning_rate": 3.081443979028098e-07,
      "loss": 1.0512,
      "step": 5077
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.15387964248657227,
      "learning_rate": 3.032750904169546e-07,
      "loss": 1.1401,
      "step": 5078
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1410612016916275,
      "learning_rate": 2.9844450399024726e-07,
      "loss": 1.1895,
      "step": 5079
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.16066254675388336,
      "learning_rate": 2.9365264049884267e-07,
      "loss": 0.996,
      "step": 5080
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.15137454867362976,
      "learning_rate": 2.8889950180382986e-07,
      "loss": 1.1004,
      "step": 5081
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12712113559246063,
      "learning_rate": 2.8418508975127654e-07,
      "loss": 1.129,
      "step": 5082
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.11996760964393616,
      "learning_rate": 2.795094061722181e-07,
      "loss": 1.1293,
      "step": 5083
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.11692588031291962,
      "learning_rate": 2.7487245288261296e-07,
      "loss": 1.1384,
      "step": 5084
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1222895160317421,
      "learning_rate": 2.702742316834206e-07,
      "loss": 1.0696,
      "step": 5085
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1586683690547943,
      "learning_rate": 2.657147443605457e-07,
      "loss": 1.0832,
      "step": 5086
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12824463844299316,
      "learning_rate": 2.6119399268480507e-07,
      "loss": 1.1153,
      "step": 5087
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.13615106046199799,
      "learning_rate": 2.5671197841203865e-07,
      "loss": 1.1406,
      "step": 5088
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.19746580719947815,
      "learning_rate": 2.5226870328299845e-07,
      "loss": 1.1786,
      "step": 5089
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.14522241055965424,
      "learning_rate": 2.478641690233929e-07,
      "loss": 1.1836,
      "step": 5090
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1465136706829071,
      "learning_rate": 2.434983773439092e-07,
      "loss": 1.1937,
      "step": 5091
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1597975641489029,
      "learning_rate": 2.3917132994016877e-07,
      "loss": 1.0994,
      "step": 5092
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1316349357366562,
      "learning_rate": 2.3488302849272724e-07,
      "loss": 1.0787,
      "step": 5093
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1232524961233139,
      "learning_rate": 2.3063347466713013e-07,
      "loss": 0.9569,
      "step": 5094
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.16637974977493286,
      "learning_rate": 2.264226701138461e-07,
      "loss": 1.0339,
      "step": 5095
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1336808055639267,
      "learning_rate": 2.2225061646830025e-07,
      "loss": 1.0725,
      "step": 5096
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12152180820703506,
      "learning_rate": 2.181173153508853e-07,
      "loss": 1.0328,
      "step": 5097
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1539098024368286,
      "learning_rate": 2.1402276836691714e-07,
      "loss": 1.1397,
      "step": 5098
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12411568313837051,
      "learning_rate": 2.0996697710666812e-07,
      "loss": 1.0323,
      "step": 5099
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12702935934066772,
      "learning_rate": 2.0594994314536707e-07,
      "loss": 1.0595,
      "step": 5100
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.13131678104400635,
      "learning_rate": 2.0197166804317712e-07,
      "loss": 1.0887,
      "step": 5101
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.11571736633777618,
      "learning_rate": 1.9803215334522896e-07,
      "loss": 1.0635,
      "step": 5102
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.10822071135044098,
      "learning_rate": 1.941314005815653e-07,
      "loss": 1.1572,
      "step": 5103
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.11526574194431305,
      "learning_rate": 1.9026941126721875e-07,
      "loss": 1.1078,
      "step": 5104
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.13573724031448364,
      "learning_rate": 1.8644618690211168e-07,
      "loss": 1.1357,
      "step": 5105
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1445101648569107,
      "learning_rate": 1.826617289711563e-07,
      "loss": 1.1152,
      "step": 5106
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.13478633761405945,
      "learning_rate": 1.7891603894418796e-07,
      "loss": 1.13,
      "step": 5107
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.15783432126045227,
      "learning_rate": 1.7520911827598742e-07,
      "loss": 1.193,
      "step": 5108
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.13923774659633636,
      "learning_rate": 1.7154096840629185e-07,
      "loss": 1.2076,
      "step": 5109
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.14760057628154755,
      "learning_rate": 1.679115907597617e-07,
      "loss": 1.1971,
      "step": 5110
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.13316386938095093,
      "learning_rate": 1.643209867460027e-07,
      "loss": 1.0967,
      "step": 5111
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.13875122368335724,
      "learning_rate": 1.6076915775956604e-07,
      "loss": 1.1913,
      "step": 5112
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12592041492462158,
      "learning_rate": 1.5725610517994815e-07,
      "loss": 1.0758,
      "step": 5113
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1394103616476059,
      "learning_rate": 1.537818303715688e-07,
      "loss": 1.1775,
      "step": 5114
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.15702056884765625,
      "learning_rate": 1.50346334683793e-07,
      "loss": 1.0855,
      "step": 5115
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.11635878682136536,
      "learning_rate": 1.4694961945093122e-07,
      "loss": 1.2761,
      "step": 5116
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.13720214366912842,
      "learning_rate": 1.4359168599223928e-07,
      "loss": 1.1304,
      "step": 5117
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.13830719888210297,
      "learning_rate": 1.4027253561188502e-07,
      "loss": 1.1507,
      "step": 5118
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.18242749571800232,
      "learning_rate": 1.3699216959899287e-07,
      "loss": 1.0858,
      "step": 5119
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1335003823041916,
      "learning_rate": 1.3375058922763252e-07,
      "loss": 1.0742,
      "step": 5120
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.17322947084903717,
      "learning_rate": 1.3054779575677466e-07,
      "loss": 1.1462,
      "step": 5121
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.13940775394439697,
      "learning_rate": 1.2738379043035764e-07,
      "loss": 1.1045,
      "step": 5122
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.13278940320014954,
      "learning_rate": 1.2425857447725398e-07,
      "loss": 1.1594,
      "step": 5123
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12604530155658722,
      "learning_rate": 1.211721491112372e-07,
      "loss": 1.0794,
      "step": 5124
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12561942636966705,
      "learning_rate": 1.1812451553107062e-07,
      "loss": 1.1523,
      "step": 5125
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.13286955654621124,
      "learning_rate": 1.1511567492038522e-07,
      "loss": 1.0489,
      "step": 5126
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.13828232884407043,
      "learning_rate": 1.1214562844781285e-07,
      "loss": 1.1708,
      "step": 5127
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.13223105669021606,
      "learning_rate": 1.0921437726686411e-07,
      "loss": 1.1161,
      "step": 5128
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.11063012480735779,
      "learning_rate": 1.0632192251601725e-07,
      "loss": 1.0862,
      "step": 5129
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.2066953331232071,
      "learning_rate": 1.0346826531865139e-07,
      "loss": 1.1936,
      "step": 5130
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1320384442806244,
      "learning_rate": 1.006534067831022e-07,
      "loss": 1.0494,
      "step": 5131
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.13440391421318054,
      "learning_rate": 9.78773480026396e-08,
      "loss": 1.098,
      "step": 5132
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1648755520582199,
      "learning_rate": 9.514009005543445e-08,
      "loss": 1.1506,
      "step": 5133
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.15341328084468842,
      "learning_rate": 9.244163400462525e-08,
      "loss": 1.0949,
      "step": 5134
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.13251441717147827,
      "learning_rate": 8.978198089824031e-08,
      "loss": 1.0651,
      "step": 5135
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1273740530014038,
      "learning_rate": 8.716113176927554e-08,
      "loss": 1.1909,
      "step": 5136
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12612281739711761,
      "learning_rate": 8.457908763562783e-08,
      "loss": 1.0569,
      "step": 5137
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.10289904475212097,
      "learning_rate": 8.203584950013943e-08,
      "loss": 1.1528,
      "step": 5138
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.13712230324745178,
      "learning_rate": 7.953141835057576e-08,
      "loss": 1.1194,
      "step": 5139
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.11517661809921265,
      "learning_rate": 7.706579515962542e-08,
      "loss": 1.1475,
      "step": 5140
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12913502752780914,
      "learning_rate": 7.463898088490018e-08,
      "loss": 1.1715,
      "step": 5141
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.11220915615558624,
      "learning_rate": 7.225097646895718e-08,
      "loss": 1.1704,
      "step": 5142
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.16433511674404144,
      "learning_rate": 6.990178283927674e-08,
      "loss": 1.1535,
      "step": 5143
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.16862736642360687,
      "learning_rate": 6.759140090824013e-08,
      "loss": 1.0621,
      "step": 5144
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1357283741235733,
      "learning_rate": 6.53198315731851e-08,
      "loss": 1.1179,
      "step": 5145
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.10405337065458298,
      "learning_rate": 6.308707571636153e-08,
      "loss": 1.1235,
      "step": 5146
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.15464085340499878,
      "learning_rate": 6.089313420494236e-08,
      "loss": 1.1033,
      "step": 5147
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.14480306208133698,
      "learning_rate": 5.87380078910349e-08,
      "loss": 1.0637,
      "step": 5148
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1122102364897728,
      "learning_rate": 5.6621697611658474e-08,
      "loss": 1.0709,
      "step": 5149
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.15667085349559784,
      "learning_rate": 5.454420418877782e-08,
      "loss": 1.0028,
      "step": 5150
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12195438891649246,
      "learning_rate": 5.2505528429247494e-08,
      "loss": 1.2031,
      "step": 5151
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1484317034482956,
      "learning_rate": 5.0505671124878584e-08,
      "loss": 1.1849,
      "step": 5152
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12736758589744568,
      "learning_rate": 4.8544633052383106e-08,
      "loss": 1.1402,
      "step": 5153
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1456436961889267,
      "learning_rate": 4.662241497341846e-08,
      "loss": 1.0899,
      "step": 5154
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12857754528522491,
      "learning_rate": 4.473901763454302e-08,
      "loss": 1.1684,
      "step": 5155
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1249743327498436,
      "learning_rate": 4.289444176724944e-08,
      "loss": 1.1244,
      "step": 5156
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.17325441539287567,
      "learning_rate": 4.108868808794242e-08,
      "loss": 1.0834,
      "step": 5157
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1365194022655487,
      "learning_rate": 3.932175729797205e-08,
      "loss": 1.1266,
      "step": 5158
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.16878347098827362,
      "learning_rate": 3.759365008357829e-08,
      "loss": 1.151,
      "step": 5159
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12318456172943115,
      "learning_rate": 3.590436711594647e-08,
      "loss": 1.1102,
      "step": 5160
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1376642882823944,
      "learning_rate": 3.425390905117398e-08,
      "loss": 1.1254,
      "step": 5161
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.14057637751102448,
      "learning_rate": 3.26422765302814e-08,
      "loss": 1.1388,
      "step": 5162
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.10714004933834076,
      "learning_rate": 3.1069470179201364e-08,
      "loss": 1.1037,
      "step": 5163
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.18718230724334717,
      "learning_rate": 2.9535490608789684e-08,
      "loss": 1.0768,
      "step": 5164
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.137231707572937,
      "learning_rate": 2.8040338414847545e-08,
      "loss": 1.1609,
      "step": 5165
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.16819770634174347,
      "learning_rate": 2.6584014178054895e-08,
      "loss": 1.0976,
      "step": 5166
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.10904133319854736,
      "learning_rate": 2.516651846403706e-08,
      "loss": 1.099,
      "step": 5167
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1564498245716095,
      "learning_rate": 2.378785182333143e-08,
      "loss": 1.1847,
      "step": 5168
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.12908276915550232,
      "learning_rate": 2.2448014791398574e-08,
      "loss": 1.1867,
      "step": 5169
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.13889434933662415,
      "learning_rate": 2.1147007888622228e-08,
      "loss": 1.1094,
      "step": 5170
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.10027969628572464,
      "learning_rate": 1.9884831620287093e-08,
      "loss": 1.1149,
      "step": 5171
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.12295103818178177,
      "learning_rate": 1.8661486476612145e-08,
      "loss": 1.1355,
      "step": 5172
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.12101224064826965,
      "learning_rate": 1.747697293272843e-08,
      "loss": 1.0102,
      "step": 5173
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.14703305065631866,
      "learning_rate": 1.633129144870127e-08,
      "loss": 1.097,
      "step": 5174
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.14182446897029877,
      "learning_rate": 1.522444246947474e-08,
      "loss": 1.2005,
      "step": 5175
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.13692399859428406,
      "learning_rate": 1.4156426424960512e-08,
      "loss": 1.198,
      "step": 5176
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.15482723712921143,
      "learning_rate": 1.3127243729949001e-08,
      "loss": 1.0958,
      "step": 5177
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.12303395569324493,
      "learning_rate": 1.2136894784176011e-08,
      "loss": 1.0791,
      "step": 5178
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.13226960599422455,
      "learning_rate": 1.1185379972256105e-08,
      "loss": 1.1302,
      "step": 5179
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.12948869168758392,
      "learning_rate": 1.0272699663782526e-08,
      "loss": 1.1276,
      "step": 5180
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.12011745572090149,
      "learning_rate": 9.398854213193974e-09,
      "loss": 1.1504,
      "step": 5181
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.11959274113178253,
      "learning_rate": 8.563843959907836e-09,
      "loss": 1.1797,
      "step": 5182
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1451667696237564,
      "learning_rate": 7.767669228231356e-09,
      "loss": 1.0928,
      "step": 5183
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.11362451314926147,
      "learning_rate": 7.01033032737275e-09,
      "loss": 1.1211,
      "step": 5184
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1453721821308136,
      "learning_rate": 6.291827551474505e-09,
      "loss": 1.0867,
      "step": 5185
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.19437313079833984,
      "learning_rate": 5.6121611796133845e-09,
      "loss": 1.0596,
      "step": 5186
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1342795491218567,
      "learning_rate": 4.971331475756014e-09,
      "loss": 1.1208,
      "step": 5187
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.13887731730937958,
      "learning_rate": 4.369338688781088e-09,
      "loss": 1.1218,
      "step": 5188
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1296427994966507,
      "learning_rate": 3.806183052512679e-09,
      "loss": 1.0877,
      "step": 5189
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.10996977239847183,
      "learning_rate": 3.2818647856647234e-09,
      "loss": 1.2789,
      "step": 5190
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.12941065430641174,
      "learning_rate": 2.7963840918854288e-09,
      "loss": 1.131,
      "step": 5191
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.13142667710781097,
      "learning_rate": 2.3497411597128703e-09,
      "loss": 1.2408,
      "step": 5192
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1832188367843628,
      "learning_rate": 1.9419361626416e-09,
      "loss": 1.1264,
      "step": 5193
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1453620344400406,
      "learning_rate": 1.5729692590338297e-09,
      "loss": 1.0964,
      "step": 5194
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.16202779114246368,
      "learning_rate": 1.2428405921971476e-09,
      "loss": 1.1125,
      "step": 5195
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1421562284231186,
      "learning_rate": 9.515502903734153e-10,
      "loss": 1.108,
      "step": 5196
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.11299877613782883,
      "learning_rate": 6.990984666610523e-10,
      "loss": 1.1447,
      "step": 5197
    },
    {
      "epoch": 1.0,
      "step": 5197,
      "total_flos": 8.28445937327145e+18,
      "train_loss": 1.1527900072802069,
      "train_runtime": 30836.1705,
      "train_samples_per_second": 21.575,
      "train_steps_per_second": 0.169
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 5197,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 8.28445937327145e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}