{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 46.15384615384615,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015384615384615385,
      "grad_norm": 5.84375,
      "learning_rate": 0.0,
      "loss": 0.7089,
      "step": 1
    },
    {
      "epoch": 0.03076923076923077,
      "grad_norm": 3.984375,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 0.731,
      "step": 2
    },
    {
      "epoch": 0.046153846153846156,
      "grad_norm": 3.84375,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 0.7307,
      "step": 3
    },
    {
      "epoch": 0.06153846153846154,
      "grad_norm": 7.71875,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 0.7422,
      "step": 4
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 3.65625,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 0.7331,
      "step": 5
    },
    {
      "epoch": 0.09230769230769231,
      "grad_norm": 8.1875,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.7315,
      "step": 6
    },
    {
      "epoch": 0.1076923076923077,
      "grad_norm": 12.4375,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.7384,
      "step": 7
    },
    {
      "epoch": 0.12307692307692308,
      "grad_norm": 20.625,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 0.7246,
      "step": 8
    },
    {
      "epoch": 0.13846153846153847,
      "grad_norm": 16.25,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.7433,
      "step": 9
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 3.3125,
      "learning_rate": 4.5e-07,
      "loss": 0.711,
      "step": 10
    },
    {
      "epoch": 0.16923076923076924,
      "grad_norm": 12.5625,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.7272,
      "step": 11
    },
    {
      "epoch": 0.18461538461538463,
      "grad_norm": 7.3125,
      "learning_rate": 5.5e-07,
      "loss": 0.723,
      "step": 12
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.28125,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.7242,
      "step": 13
    },
    {
      "epoch": 0.2153846153846154,
      "grad_norm": 3.875,
      "learning_rate": 6.5e-07,
      "loss": 0.7276,
      "step": 14
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 13.75,
      "learning_rate": 7.000000000000001e-07,
      "loss": 0.7266,
      "step": 15
    },
    {
      "epoch": 0.24615384615384617,
      "grad_norm": 9.4375,
      "learning_rate": 7.5e-07,
      "loss": 0.7106,
      "step": 16
    },
    {
      "epoch": 0.26153846153846155,
      "grad_norm": 8.1875,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.7199,
      "step": 17
    },
    {
      "epoch": 0.27692307692307694,
      "grad_norm": 14.0625,
      "learning_rate": 8.500000000000001e-07,
      "loss": 0.7315,
      "step": 18
    },
    {
      "epoch": 0.2923076923076923,
      "grad_norm": 3.453125,
      "learning_rate": 9e-07,
      "loss": 0.7139,
      "step": 19
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 5.96875,
      "learning_rate": 9.5e-07,
      "loss": 0.7279,
      "step": 20
    },
    {
      "epoch": 0.3230769230769231,
      "grad_norm": 3.59375,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.7326,
      "step": 21
    },
    {
      "epoch": 0.3384615384615385,
      "grad_norm": 6.5625,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 0.7166,
      "step": 22
    },
    {
      "epoch": 0.35384615384615387,
      "grad_norm": 5.59375,
      "learning_rate": 1.1e-06,
      "loss": 0.705,
      "step": 23
    },
    {
      "epoch": 0.36923076923076925,
      "grad_norm": 14.375,
      "learning_rate": 1.15e-06,
      "loss": 0.7232,
      "step": 24
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 5.21875,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.7176,
      "step": 25
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.3125,
      "learning_rate": 1.25e-06,
      "loss": 0.7209,
      "step": 26
    },
    {
      "epoch": 0.4153846153846154,
      "grad_norm": 6.46875,
      "learning_rate": 1.3e-06,
      "loss": 0.7165,
      "step": 27
    },
    {
      "epoch": 0.4307692307692308,
      "grad_norm": 9.625,
      "learning_rate": 1.35e-06,
      "loss": 0.7282,
      "step": 28
    },
    {
      "epoch": 0.4461538461538462,
      "grad_norm": 17.75,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.7182,
      "step": 29
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 4.46875,
      "learning_rate": 1.45e-06,
      "loss": 0.7108,
      "step": 30
    },
    {
      "epoch": 0.47692307692307695,
      "grad_norm": 7.0,
      "learning_rate": 1.5e-06,
      "loss": 0.7216,
      "step": 31
    },
    {
      "epoch": 0.49230769230769234,
      "grad_norm": 4.625,
      "learning_rate": 1.55e-06,
      "loss": 0.7194,
      "step": 32
    },
    {
      "epoch": 0.5076923076923077,
      "grad_norm": 7.71875,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.7294,
      "step": 33
    },
    {
      "epoch": 0.5230769230769231,
      "grad_norm": 7.84375,
      "learning_rate": 1.65e-06,
      "loss": 0.7147,
      "step": 34
    },
    {
      "epoch": 0.5384615384615384,
      "grad_norm": 12.5625,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.724,
      "step": 35
    },
    {
      "epoch": 0.5538461538461539,
      "grad_norm": 8.3125,
      "learning_rate": 1.7500000000000002e-06,
      "loss": 0.7173,
      "step": 36
    },
    {
      "epoch": 0.5692307692307692,
      "grad_norm": 4.5625,
      "learning_rate": 1.8e-06,
      "loss": 0.7181,
      "step": 37
    },
    {
      "epoch": 0.5846153846153846,
      "grad_norm": 5.0625,
      "learning_rate": 1.85e-06,
      "loss": 0.715,
      "step": 38
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.96875,
      "learning_rate": 1.9e-06,
      "loss": 0.7197,
      "step": 39
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 3.484375,
      "learning_rate": 1.95e-06,
      "loss": 0.7196,
      "step": 40
    },
    {
      "epoch": 0.6307692307692307,
      "grad_norm": 10.4375,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.7168,
      "step": 41
    },
    {
      "epoch": 0.6461538461538462,
      "grad_norm": 12.0,
      "learning_rate": 2.0500000000000003e-06,
      "loss": 0.7096,
      "step": 42
    },
    {
      "epoch": 0.6615384615384615,
      "grad_norm": 12.25,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.7265,
      "step": 43
    },
    {
      "epoch": 0.676923076923077,
      "grad_norm": 4.5625,
      "learning_rate": 2.1499999999999997e-06,
      "loss": 0.7216,
      "step": 44
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 4.6875,
      "learning_rate": 2.2e-06,
      "loss": 0.7124,
      "step": 45
    },
    {
      "epoch": 0.7076923076923077,
      "grad_norm": 20.875,
      "learning_rate": 2.25e-06,
      "loss": 0.712,
      "step": 46
    },
    {
      "epoch": 0.7230769230769231,
      "grad_norm": 7.78125,
      "learning_rate": 2.3e-06,
      "loss": 0.7162,
      "step": 47
    },
    {
      "epoch": 0.7384615384615385,
      "grad_norm": 6.75,
      "learning_rate": 2.35e-06,
      "loss": 0.7057,
      "step": 48
    },
    {
      "epoch": 0.7538461538461538,
      "grad_norm": 5.78125,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.7059,
      "step": 49
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 11.875,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.7117,
      "step": 50
    },
    {
      "epoch": 0.7846153846153846,
      "grad_norm": 3.671875,
      "learning_rate": 2.5e-06,
      "loss": 0.7089,
      "step": 51
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.0,
      "learning_rate": 2.55e-06,
      "loss": 0.7127,
      "step": 52
    },
    {
      "epoch": 0.8153846153846154,
      "grad_norm": 23.5,
      "learning_rate": 2.6e-06,
      "loss": 0.7251,
      "step": 53
    },
    {
      "epoch": 0.8307692307692308,
      "grad_norm": 22.0,
      "learning_rate": 2.65e-06,
      "loss": 0.7251,
      "step": 54
    },
    {
      "epoch": 0.8461538461538461,
      "grad_norm": 3.640625,
      "learning_rate": 2.7e-06,
      "loss": 0.7064,
      "step": 55
    },
    {
      "epoch": 0.8615384615384616,
      "grad_norm": 6.59375,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.6976,
      "step": 56
    },
    {
      "epoch": 0.8769230769230769,
      "grad_norm": 5.21875,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.7091,
      "step": 57
    },
    {
      "epoch": 0.8923076923076924,
      "grad_norm": 9.625,
      "learning_rate": 2.8500000000000002e-06,
      "loss": 0.7091,
      "step": 58
    },
    {
      "epoch": 0.9076923076923077,
      "grad_norm": 15.625,
      "learning_rate": 2.9e-06,
      "loss": 0.7043,
      "step": 59
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 7.65625,
      "learning_rate": 2.95e-06,
      "loss": 0.7096,
      "step": 60
    },
    {
      "epoch": 0.9384615384615385,
      "grad_norm": 5.09375,
      "learning_rate": 3e-06,
      "loss": 0.6966,
      "step": 61
    },
    {
      "epoch": 0.9538461538461539,
      "grad_norm": 7.78125,
      "learning_rate": 3.05e-06,
      "loss": 0.6932,
      "step": 62
    },
    {
      "epoch": 0.9692307692307692,
      "grad_norm": 18.25,
      "learning_rate": 3.1e-06,
      "loss": 0.7129,
      "step": 63
    },
    {
      "epoch": 0.9846153846153847,
      "grad_norm": 11.8125,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.7161,
      "step": 64
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.125,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.7019,
      "step": 65
    },
    {
      "epoch": 1.0153846153846153,
      "grad_norm": 12.3125,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.7061,
      "step": 66
    },
    {
      "epoch": 1.0307692307692307,
      "grad_norm": 5.46875,
      "learning_rate": 3.3e-06,
      "loss": 0.7028,
      "step": 67
    },
    {
      "epoch": 1.0461538461538462,
      "grad_norm": 10.0625,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.7029,
      "step": 68
    },
    {
      "epoch": 1.0615384615384615,
      "grad_norm": 6.75,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.7078,
      "step": 69
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 17.625,
      "learning_rate": 3.4500000000000004e-06,
      "loss": 0.7223,
      "step": 70
    },
    {
      "epoch": 1.0923076923076924,
      "grad_norm": 23.25,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 0.7106,
      "step": 71
    },
    {
      "epoch": 1.1076923076923078,
      "grad_norm": 6.6875,
      "learning_rate": 3.55e-06,
      "loss": 0.6821,
      "step": 72
    },
    {
      "epoch": 1.123076923076923,
      "grad_norm": 7.0625,
      "learning_rate": 3.6e-06,
      "loss": 0.6978,
      "step": 73
    },
    {
      "epoch": 1.1384615384615384,
      "grad_norm": 11.9375,
      "learning_rate": 3.6499999999999998e-06,
      "loss": 0.6941,
      "step": 74
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 11.25,
      "learning_rate": 3.7e-06,
      "loss": 0.685,
      "step": 75
    },
    {
      "epoch": 1.1692307692307693,
      "grad_norm": 21.375,
      "learning_rate": 3.75e-06,
      "loss": 0.6735,
      "step": 76
    },
    {
      "epoch": 1.1846153846153846,
      "grad_norm": 17.375,
      "learning_rate": 3.8e-06,
      "loss": 0.6894,
      "step": 77
    },
    {
      "epoch": 1.2,
      "grad_norm": 17.75,
      "learning_rate": 3.85e-06,
      "loss": 0.6708,
      "step": 78
    },
    {
      "epoch": 1.2153846153846155,
      "grad_norm": 21.25,
      "learning_rate": 3.9e-06,
      "loss": 0.6759,
      "step": 79
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 26.125,
      "learning_rate": 3.95e-06,
      "loss": 0.6756,
      "step": 80
    },
    {
      "epoch": 1.2461538461538462,
      "grad_norm": 19.375,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.6529,
      "step": 81
    },
    {
      "epoch": 1.2615384615384615,
      "grad_norm": 15.8125,
      "learning_rate": 4.05e-06,
      "loss": 0.6216,
      "step": 82
    },
    {
      "epoch": 1.2769230769230768,
      "grad_norm": 44.0,
      "learning_rate": 4.1000000000000006e-06,
      "loss": 0.6285,
      "step": 83
    },
    {
      "epoch": 1.2923076923076924,
      "grad_norm": 36.25,
      "learning_rate": 4.15e-06,
      "loss": 0.6371,
      "step": 84
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 41.25,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.5986,
      "step": 85
    },
    {
      "epoch": 1.323076923076923,
      "grad_norm": 106.0,
      "learning_rate": 4.250000000000001e-06,
      "loss": 0.6443,
      "step": 86
    },
    {
      "epoch": 1.3384615384615386,
      "grad_norm": 39.25,
      "learning_rate": 4.2999999999999995e-06,
      "loss": 0.5701,
      "step": 87
    },
    {
      "epoch": 1.353846153846154,
      "grad_norm": 17.5,
      "learning_rate": 4.35e-06,
      "loss": 0.5921,
      "step": 88
    },
    {
      "epoch": 1.3692307692307693,
      "grad_norm": 103.5,
      "learning_rate": 4.4e-06,
      "loss": 0.6055,
      "step": 89
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 62.5,
      "learning_rate": 4.45e-06,
      "loss": 0.6721,
      "step": 90
    },
    {
      "epoch": 1.4,
      "grad_norm": 41.5,
      "learning_rate": 4.5e-06,
      "loss": 0.6112,
      "step": 91
    },
    {
      "epoch": 1.4153846153846155,
      "grad_norm": 24.0,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.5529,
      "step": 92
    },
    {
      "epoch": 1.4307692307692308,
      "grad_norm": 36.5,
      "learning_rate": 4.6e-06,
      "loss": 0.5706,
      "step": 93
    },
    {
      "epoch": 1.4461538461538461,
      "grad_norm": 72.0,
      "learning_rate": 4.65e-06,
      "loss": 0.5816,
      "step": 94
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 64.5,
      "learning_rate": 4.7e-06,
      "loss": 0.6636,
      "step": 95
    },
    {
      "epoch": 1.476923076923077,
      "grad_norm": 61.0,
      "learning_rate": 4.75e-06,
      "loss": 0.6361,
      "step": 96
    },
    {
      "epoch": 1.4923076923076923,
      "grad_norm": 30.75,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.4694,
      "step": 97
    },
    {
      "epoch": 1.5076923076923077,
      "grad_norm": 41.75,
      "learning_rate": 4.85e-06,
      "loss": 0.5414,
      "step": 98
    },
    {
      "epoch": 1.523076923076923,
      "grad_norm": 26.25,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.5588,
      "step": 99
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 81.5,
      "learning_rate": 4.950000000000001e-06,
      "loss": 0.5768,
      "step": 100
    },
    {
      "epoch": 1.5538461538461539,
      "grad_norm": 66.0,
      "learning_rate": 5e-06,
      "loss": 0.6162,
      "step": 101
    },
    {
      "epoch": 1.5692307692307692,
      "grad_norm": 72.5,
      "learning_rate": 5.050000000000001e-06,
      "loss": 0.5524,
      "step": 102
    },
    {
      "epoch": 1.5846153846153848,
      "grad_norm": 39.25,
      "learning_rate": 5.1e-06,
      "loss": 0.5442,
      "step": 103
    },
    {
      "epoch": 1.6,
      "grad_norm": 25.75,
      "learning_rate": 5.15e-06,
      "loss": 0.4873,
      "step": 104
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 37.0,
      "learning_rate": 5.2e-06,
      "loss": 0.4799,
      "step": 105
    },
    {
      "epoch": 1.6307692307692307,
      "grad_norm": 21.75,
      "learning_rate": 5.25e-06,
      "loss": 0.4769,
      "step": 106
    },
    {
      "epoch": 1.646153846153846,
      "grad_norm": 34.75,
      "learning_rate": 5.3e-06,
      "loss": 0.5045,
      "step": 107
    },
    {
      "epoch": 1.6615384615384614,
      "grad_norm": 30.125,
      "learning_rate": 5.3500000000000004e-06,
      "loss": 0.4787,
      "step": 108
    },
    {
      "epoch": 1.676923076923077,
      "grad_norm": 34.0,
      "learning_rate": 5.4e-06,
      "loss": 0.4373,
      "step": 109
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 40.5,
      "learning_rate": 5.45e-06,
      "loss": 0.5179,
      "step": 110
    },
    {
      "epoch": 1.7076923076923078,
      "grad_norm": 31.75,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.5061,
      "step": 111
    },
    {
      "epoch": 1.7230769230769232,
      "grad_norm": 49.5,
      "learning_rate": 5.55e-06,
      "loss": 0.4606,
      "step": 112
    },
    {
      "epoch": 1.7384615384615385,
      "grad_norm": 54.25,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.4305,
      "step": 113
    },
    {
      "epoch": 1.7538461538461538,
      "grad_norm": 48.75,
      "learning_rate": 5.65e-06,
      "loss": 0.4597,
      "step": 114
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 99.5,
      "learning_rate": 5.7000000000000005e-06,
      "loss": 0.4762,
      "step": 115
    },
    {
      "epoch": 1.7846153846153845,
      "grad_norm": 128.0,
      "learning_rate": 5.750000000000001e-06,
      "loss": 0.4943,
      "step": 116
    },
    {
      "epoch": 1.8,
      "grad_norm": 111.0,
      "learning_rate": 5.8e-06,
      "loss": 0.4417,
      "step": 117
    },
    {
      "epoch": 1.8153846153846154,
      "grad_norm": 108.0,
      "learning_rate": 5.850000000000001e-06,
      "loss": 0.5368,
      "step": 118
    },
    {
      "epoch": 1.830769230769231,
      "grad_norm": 48.5,
      "learning_rate": 5.9e-06,
      "loss": 0.5244,
      "step": 119
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 28.375,
      "learning_rate": 5.95e-06,
      "loss": 0.451,
      "step": 120
    },
    {
      "epoch": 1.8615384615384616,
      "grad_norm": 34.5,
      "learning_rate": 6e-06,
      "loss": 0.3586,
      "step": 121
    },
    {
      "epoch": 1.876923076923077,
      "grad_norm": 64.0,
      "learning_rate": 6.0500000000000005e-06,
      "loss": 0.4117,
      "step": 122
    },
    {
      "epoch": 1.8923076923076922,
      "grad_norm": 69.5,
      "learning_rate": 6.1e-06,
      "loss": 0.415,
      "step": 123
    },
    {
      "epoch": 1.9076923076923076,
      "grad_norm": 31.375,
      "learning_rate": 6.15e-06,
      "loss": 0.371,
      "step": 124
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 70.0,
      "learning_rate": 6.2e-06,
      "loss": 0.4339,
      "step": 125
    },
    {
      "epoch": 1.9384615384615385,
      "grad_norm": 29.75,
      "learning_rate": 6.25e-06,
      "loss": 0.341,
      "step": 126
    },
    {
      "epoch": 1.953846153846154,
      "grad_norm": 25.375,
      "learning_rate": 6.300000000000001e-06,
      "loss": 0.35,
      "step": 127
    },
    {
      "epoch": 1.9692307692307693,
      "grad_norm": 63.5,
      "learning_rate": 6.35e-06,
      "loss": 0.3487,
      "step": 128
    },
    {
      "epoch": 1.9846153846153847,
      "grad_norm": 23.75,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.4917,
      "step": 129
    },
    {
      "epoch": 2.0,
      "grad_norm": 39.0,
      "learning_rate": 6.45e-06,
      "loss": 0.4082,
      "step": 130
    },
    {
      "epoch": 2.0153846153846153,
      "grad_norm": 48.25,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.3158,
      "step": 131
    },
    {
      "epoch": 2.0307692307692307,
      "grad_norm": 41.0,
      "learning_rate": 6.550000000000001e-06,
      "loss": 0.3175,
      "step": 132
    },
    {
      "epoch": 2.046153846153846,
      "grad_norm": 51.25,
      "learning_rate": 6.6e-06,
      "loss": 0.2999,
      "step": 133
    },
    {
      "epoch": 2.0615384615384613,
      "grad_norm": 37.0,
      "learning_rate": 6.650000000000001e-06,
      "loss": 0.405,
      "step": 134
    },
    {
      "epoch": 2.076923076923077,
      "grad_norm": 19.0,
      "learning_rate": 6.700000000000001e-06,
      "loss": 0.3028,
      "step": 135
    },
    {
      "epoch": 2.0923076923076924,
      "grad_norm": 53.25,
      "learning_rate": 6.750000000000001e-06,
      "loss": 0.3518,
      "step": 136
    },
    {
      "epoch": 2.1076923076923078,
      "grad_norm": 54.25,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.2722,
      "step": 137
    },
    {
      "epoch": 2.123076923076923,
      "grad_norm": 39.25,
      "learning_rate": 6.8500000000000005e-06,
      "loss": 0.3335,
      "step": 138
    },
    {
      "epoch": 2.1384615384615384,
      "grad_norm": 28.625,
      "learning_rate": 6.900000000000001e-06,
      "loss": 0.3085,
      "step": 139
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 41.0,
      "learning_rate": 6.950000000000001e-06,
      "loss": 0.2804,
      "step": 140
    },
    {
      "epoch": 2.169230769230769,
      "grad_norm": 48.0,
      "learning_rate": 7.000000000000001e-06,
      "loss": 0.3294,
      "step": 141
    },
    {
      "epoch": 2.184615384615385,
      "grad_norm": 31.875,
      "learning_rate": 7.049999999999999e-06,
      "loss": 0.3626,
      "step": 142
    },
    {
      "epoch": 2.2,
      "grad_norm": 37.5,
      "learning_rate": 7.1e-06,
      "loss": 0.325,
      "step": 143
    },
    {
      "epoch": 2.2153846153846155,
      "grad_norm": 114.0,
      "learning_rate": 7.15e-06,
      "loss": 0.3128,
      "step": 144
    },
    {
      "epoch": 2.230769230769231,
      "grad_norm": 65.5,
      "learning_rate": 7.2e-06,
      "loss": 0.2666,
      "step": 145
    },
    {
      "epoch": 2.246153846153846,
      "grad_norm": 31.5,
      "learning_rate": 7.25e-06,
      "loss": 0.3073,
      "step": 146
    },
    {
      "epoch": 2.2615384615384615,
      "grad_norm": 57.0,
      "learning_rate": 7.2999999999999996e-06,
      "loss": 0.3931,
      "step": 147
    },
    {
      "epoch": 2.276923076923077,
      "grad_norm": 51.5,
      "learning_rate": 7.35e-06,
      "loss": 0.286,
      "step": 148
    },
    {
      "epoch": 2.292307692307692,
      "grad_norm": 51.25,
      "learning_rate": 7.4e-06,
      "loss": 0.2188,
      "step": 149
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 106.5,
      "learning_rate": 7.45e-06,
      "loss": 0.3625,
      "step": 150
    },
    {
      "epoch": 2.3230769230769233,
      "grad_norm": 21.625,
      "learning_rate": 7.5e-06,
      "loss": 0.2163,
      "step": 151
    },
    {
      "epoch": 2.3384615384615386,
      "grad_norm": 29.125,
      "learning_rate": 7.55e-06,
      "loss": 0.2506,
      "step": 152
    },
    {
      "epoch": 2.353846153846154,
      "grad_norm": 19.625,
      "learning_rate": 7.6e-06,
      "loss": 0.1606,
      "step": 153
    },
    {
      "epoch": 2.3692307692307693,
      "grad_norm": 38.5,
      "learning_rate": 7.65e-06,
      "loss": 0.2434,
      "step": 154
    },
    {
      "epoch": 2.3846153846153846,
      "grad_norm": 50.25,
      "learning_rate": 7.7e-06,
      "loss": 0.2241,
      "step": 155
    },
    {
      "epoch": 2.4,
      "grad_norm": 56.0,
      "learning_rate": 7.75e-06,
      "loss": 0.3239,
      "step": 156
    },
    {
      "epoch": 2.4153846153846152,
      "grad_norm": 102.5,
      "learning_rate": 7.8e-06,
      "loss": 0.2467,
      "step": 157
    },
    {
      "epoch": 2.430769230769231,
      "grad_norm": 99.5,
      "learning_rate": 7.850000000000001e-06,
      "loss": 0.2407,
      "step": 158
    },
    {
      "epoch": 2.4461538461538463,
      "grad_norm": 58.75,
      "learning_rate": 7.9e-06,
      "loss": 0.2796,
      "step": 159
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 52.0,
      "learning_rate": 7.95e-06,
      "loss": 0.2519,
      "step": 160
    },
    {
      "epoch": 2.476923076923077,
      "grad_norm": 124.5,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.5565,
      "step": 161
    },
    {
      "epoch": 2.4923076923076923,
      "grad_norm": 103.0,
      "learning_rate": 8.050000000000001e-06,
      "loss": 0.3969,
      "step": 162
    },
    {
      "epoch": 2.5076923076923077,
      "grad_norm": 126.5,
      "learning_rate": 8.1e-06,
      "loss": 0.3433,
      "step": 163
    },
    {
      "epoch": 2.523076923076923,
      "grad_norm": 84.0,
      "learning_rate": 8.15e-06,
      "loss": 0.3114,
      "step": 164
    },
    {
      "epoch": 2.5384615384615383,
      "grad_norm": 16.0,
      "learning_rate": 8.200000000000001e-06,
      "loss": 0.2997,
      "step": 165
    },
    {
      "epoch": 2.5538461538461537,
      "grad_norm": 143.0,
      "learning_rate": 8.25e-06,
      "loss": 0.5248,
      "step": 166
    },
    {
      "epoch": 2.569230769230769,
      "grad_norm": 61.5,
      "learning_rate": 8.3e-06,
      "loss": 0.2721,
      "step": 167
    },
    {
      "epoch": 2.5846153846153848,
      "grad_norm": 57.25,
      "learning_rate": 8.350000000000001e-06,
      "loss": 0.2695,
      "step": 168
    },
    {
      "epoch": 2.6,
      "grad_norm": 73.0,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.4285,
      "step": 169
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 21.75,
      "learning_rate": 8.45e-06,
      "loss": 0.4448,
      "step": 170
    },
    {
      "epoch": 2.6307692307692307,
      "grad_norm": 48.75,
      "learning_rate": 8.500000000000002e-06,
      "loss": 0.3604,
      "step": 171
    },
    {
      "epoch": 2.646153846153846,
      "grad_norm": 77.0,
      "learning_rate": 8.550000000000001e-06,
      "loss": 0.4495,
      "step": 172
    },
    {
      "epoch": 2.6615384615384614,
      "grad_norm": 52.0,
      "learning_rate": 8.599999999999999e-06,
      "loss": 0.2795,
      "step": 173
    },
    {
      "epoch": 2.676923076923077,
      "grad_norm": 24.75,
      "learning_rate": 8.65e-06,
      "loss": 0.3096,
      "step": 174
    },
    {
      "epoch": 2.6923076923076925,
      "grad_norm": 18.625,
      "learning_rate": 8.7e-06,
      "loss": 0.2252,
      "step": 175
    },
    {
      "epoch": 2.707692307692308,
      "grad_norm": 23.375,
      "learning_rate": 8.75e-06,
      "loss": 0.2756,
      "step": 176
    },
    {
      "epoch": 2.723076923076923,
      "grad_norm": 67.5,
      "learning_rate": 8.8e-06,
      "loss": 0.2417,
      "step": 177
    },
    {
      "epoch": 2.7384615384615385,
      "grad_norm": 37.5,
      "learning_rate": 8.85e-06,
      "loss": 0.3493,
      "step": 178
    },
    {
      "epoch": 2.753846153846154,
      "grad_norm": 35.25,
      "learning_rate": 8.9e-06,
      "loss": 0.2727,
      "step": 179
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 21.75,
      "learning_rate": 8.95e-06,
      "loss": 0.2546,
      "step": 180
    },
    {
      "epoch": 2.7846153846153845,
      "grad_norm": 82.0,
      "learning_rate": 9e-06,
      "loss": 0.4433,
      "step": 181
    },
    {
      "epoch": 2.8,
      "grad_norm": 54.5,
      "learning_rate": 9.05e-06,
      "loss": 0.2535,
      "step": 182
    },
    {
      "epoch": 2.815384615384615,
      "grad_norm": 34.0,
      "learning_rate": 9.100000000000001e-06,
      "loss": 0.2261,
      "step": 183
    },
    {
      "epoch": 2.830769230769231,
      "grad_norm": 38.5,
      "learning_rate": 9.15e-06,
      "loss": 0.2371,
      "step": 184
    },
    {
      "epoch": 2.8461538461538463,
      "grad_norm": 25.375,
      "learning_rate": 9.2e-06,
      "loss": 0.3832,
      "step": 185
    },
    {
      "epoch": 2.8615384615384616,
      "grad_norm": 25.125,
      "learning_rate": 9.25e-06,
      "loss": 0.3851,
      "step": 186
    },
    {
      "epoch": 2.876923076923077,
      "grad_norm": 15.8125,
      "learning_rate": 9.3e-06,
      "loss": 0.1774,
      "step": 187
    },
    {
      "epoch": 2.8923076923076922,
      "grad_norm": 31.625,
      "learning_rate": 9.35e-06,
      "loss": 0.2195,
      "step": 188
    },
    {
      "epoch": 2.9076923076923076,
      "grad_norm": 32.5,
      "learning_rate": 9.4e-06,
      "loss": 0.1931,
      "step": 189
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 21.125,
      "learning_rate": 9.450000000000001e-06,
      "loss": 0.2797,
      "step": 190
    },
    {
      "epoch": 2.9384615384615387,
      "grad_norm": 26.125,
      "learning_rate": 9.5e-06,
      "loss": 0.255,
      "step": 191
    },
    {
      "epoch": 2.953846153846154,
      "grad_norm": 37.5,
      "learning_rate": 9.55e-06,
      "loss": 0.2163,
      "step": 192
    },
    {
      "epoch": 2.9692307692307693,
      "grad_norm": 26.0,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.2017,
      "step": 193
    },
    {
      "epoch": 2.9846153846153847,
      "grad_norm": 24.5,
      "learning_rate": 9.65e-06,
      "loss": 0.2054,
      "step": 194
    },
    {
      "epoch": 3.0,
      "grad_norm": 29.75,
      "learning_rate": 9.7e-06,
      "loss": 0.2678,
      "step": 195
    },
    {
      "epoch": 3.0153846153846153,
      "grad_norm": 39.25,
      "learning_rate": 9.750000000000002e-06,
      "loss": 0.2489,
      "step": 196
    },
    {
      "epoch": 3.0307692307692307,
      "grad_norm": 21.875,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.1788,
      "step": 197
    },
    {
      "epoch": 3.046153846153846,
      "grad_norm": 24.125,
      "learning_rate": 9.85e-06,
      "loss": 0.2048,
      "step": 198
    },
    {
      "epoch": 3.0615384615384613,
      "grad_norm": 10.375,
      "learning_rate": 9.900000000000002e-06,
      "loss": 0.1932,
      "step": 199
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 48.75,
      "learning_rate": 9.950000000000001e-06,
      "loss": 0.2324,
      "step": 200
    },
    {
      "epoch": 3.0923076923076924,
      "grad_norm": 35.0,
      "learning_rate": 1e-05,
      "loss": 0.2348,
      "step": 201
    },
    {
      "epoch": 3.1076923076923078,
      "grad_norm": 18.5,
      "learning_rate": 1.005e-05,
      "loss": 0.1463,
      "step": 202
    },
    {
      "epoch": 3.123076923076923,
      "grad_norm": 50.75,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 0.1917,
      "step": 203
    },
    {
      "epoch": 3.1384615384615384,
      "grad_norm": 60.0,
      "learning_rate": 1.0150000000000001e-05,
      "loss": 0.2485,
      "step": 204
    },
    {
      "epoch": 3.1538461538461537,
      "grad_norm": 48.5,
      "learning_rate": 1.02e-05,
      "loss": 0.1768,
      "step": 205
    },
    {
      "epoch": 3.169230769230769,
      "grad_norm": 14.0625,
      "learning_rate": 1.025e-05,
      "loss": 0.2604,
      "step": 206
    },
    {
      "epoch": 3.184615384615385,
      "grad_norm": 56.75,
      "learning_rate": 1.03e-05,
      "loss": 0.192,
      "step": 207
    },
    {
      "epoch": 3.2,
      "grad_norm": 55.5,
      "learning_rate": 1.035e-05,
      "loss": 0.2907,
      "step": 208
    },
    {
      "epoch": 3.2153846153846155,
      "grad_norm": 28.125,
      "learning_rate": 1.04e-05,
      "loss": 0.2862,
      "step": 209
    },
    {
      "epoch": 3.230769230769231,
      "grad_norm": 22.375,
      "learning_rate": 1.045e-05,
      "loss": 0.2503,
      "step": 210
    },
    {
      "epoch": 3.246153846153846,
      "grad_norm": 19.875,
      "learning_rate": 1.05e-05,
      "loss": 0.1399,
      "step": 211
    },
    {
      "epoch": 3.2615384615384615,
      "grad_norm": 46.0,
      "learning_rate": 1.055e-05,
      "loss": 0.2381,
      "step": 212
    },
    {
      "epoch": 3.276923076923077,
      "grad_norm": 17.375,
      "learning_rate": 1.06e-05,
      "loss": 0.2044,
      "step": 213
    },
    {
      "epoch": 3.292307692307692,
      "grad_norm": 20.5,
      "learning_rate": 1.065e-05,
      "loss": 0.2054,
      "step": 214
    },
    {
      "epoch": 3.3076923076923075,
      "grad_norm": 19.0,
      "learning_rate": 1.0700000000000001e-05,
      "loss": 0.2083,
      "step": 215
    },
    {
      "epoch": 3.3230769230769233,
      "grad_norm": 18.75,
      "learning_rate": 1.075e-05,
      "loss": 0.157,
      "step": 216
    },
    {
      "epoch": 3.3384615384615386,
      "grad_norm": 38.25,
      "learning_rate": 1.08e-05,
      "loss": 0.1927,
      "step": 217
    },
    {
      "epoch": 3.353846153846154,
      "grad_norm": 25.125,
      "learning_rate": 1.0850000000000001e-05,
      "loss": 0.1517,
      "step": 218
    },
    {
      "epoch": 3.3692307692307693,
      "grad_norm": 17.5,
      "learning_rate": 1.09e-05,
      "loss": 0.0864,
      "step": 219
    },
    {
      "epoch": 3.3846153846153846,
      "grad_norm": 38.5,
      "learning_rate": 1.095e-05,
      "loss": 0.2528,
      "step": 220
    },
    {
      "epoch": 3.4,
      "grad_norm": 32.5,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.2019,
      "step": 221
    },
    {
      "epoch": 3.4153846153846152,
      "grad_norm": 16.75,
      "learning_rate": 1.1050000000000001e-05,
      "loss": 0.2284,
      "step": 222
    },
    {
      "epoch": 3.430769230769231,
      "grad_norm": 50.25,
      "learning_rate": 1.11e-05,
      "loss": 0.1721,
      "step": 223
    },
    {
      "epoch": 3.4461538461538463,
      "grad_norm": 55.5,
      "learning_rate": 1.115e-05,
      "loss": 0.2487,
      "step": 224
    },
    {
      "epoch": 3.4615384615384617,
      "grad_norm": 28.125,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.23,
      "step": 225
    },
    {
      "epoch": 3.476923076923077,
      "grad_norm": 18.75,
      "learning_rate": 1.125e-05,
      "loss": 0.2047,
      "step": 226
    },
    {
      "epoch": 3.4923076923076923,
      "grad_norm": 24.875,
      "learning_rate": 1.13e-05,
      "loss": 0.1405,
      "step": 227
    },
    {
      "epoch": 3.5076923076923077,
      "grad_norm": 21.25,
      "learning_rate": 1.1350000000000001e-05,
      "loss": 0.1447,
      "step": 228
    },
    {
      "epoch": 3.523076923076923,
      "grad_norm": 18.125,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 0.2187,
      "step": 229
    },
    {
      "epoch": 3.5384615384615383,
      "grad_norm": 20.0,
      "learning_rate": 1.145e-05,
      "loss": 0.1467,
      "step": 230
    },
    {
      "epoch": 3.5538461538461537,
      "grad_norm": 24.625,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 0.2181,
      "step": 231
    },
    {
      "epoch": 3.569230769230769,
      "grad_norm": 31.0,
      "learning_rate": 1.1550000000000001e-05,
      "loss": 0.237,
      "step": 232
    },
    {
      "epoch": 3.5846153846153848,
      "grad_norm": 24.25,
      "learning_rate": 1.16e-05,
      "loss": 0.1976,
      "step": 233
    },
    {
      "epoch": 3.6,
      "grad_norm": 12.9375,
      "learning_rate": 1.1650000000000002e-05,
      "loss": 0.122,
      "step": 234
    },
    {
      "epoch": 3.6153846153846154,
      "grad_norm": 14.0,
      "learning_rate": 1.1700000000000001e-05,
      "loss": 0.2284,
      "step": 235
    },
    {
      "epoch": 3.6307692307692307,
      "grad_norm": 45.75,
      "learning_rate": 1.175e-05,
      "loss": 0.1597,
      "step": 236
    },
    {
      "epoch": 3.646153846153846,
      "grad_norm": 28.875,
      "learning_rate": 1.18e-05,
      "loss": 0.1511,
      "step": 237
    },
    {
      "epoch": 3.6615384615384614,
      "grad_norm": 27.25,
      "learning_rate": 1.185e-05,
      "loss": 0.2176,
      "step": 238
    },
    {
      "epoch": 3.676923076923077,
      "grad_norm": 13.8125,
      "learning_rate": 1.19e-05,
      "loss": 0.2049,
      "step": 239
    },
    {
      "epoch": 3.6923076923076925,
      "grad_norm": 15.5625,
      "learning_rate": 1.195e-05,
      "loss": 0.1225,
      "step": 240
    },
    {
      "epoch": 3.707692307692308,
      "grad_norm": 31.5,
      "learning_rate": 1.2e-05,
      "loss": 0.1787,
      "step": 241
    },
    {
      "epoch": 3.723076923076923,
      "grad_norm": 15.5625,
      "learning_rate": 1.205e-05,
      "loss": 0.2202,
      "step": 242
    },
    {
      "epoch": 3.7384615384615385,
      "grad_norm": 8.8125,
      "learning_rate": 1.2100000000000001e-05,
      "loss": 0.1514,
      "step": 243
    },
    {
      "epoch": 3.753846153846154,
      "grad_norm": 12.25,
      "learning_rate": 1.215e-05,
      "loss": 0.0959,
      "step": 244
    },
    {
      "epoch": 3.769230769230769,
      "grad_norm": 7.875,
      "learning_rate": 1.22e-05,
      "loss": 0.1655,
      "step": 245
    },
    {
      "epoch": 3.7846153846153845,
      "grad_norm": 13.75,
      "learning_rate": 1.225e-05,
      "loss": 0.1521,
      "step": 246
    },
    {
      "epoch": 3.8,
      "grad_norm": 33.75,
      "learning_rate": 1.23e-05,
      "loss": 0.3589,
      "step": 247
    },
    {
      "epoch": 3.815384615384615,
      "grad_norm": 20.0,
      "learning_rate": 1.235e-05,
      "loss": 0.1186,
      "step": 248
    },
    {
      "epoch": 3.830769230769231,
      "grad_norm": 31.875,
      "learning_rate": 1.24e-05,
      "loss": 0.2261,
      "step": 249
    },
    {
      "epoch": 3.8461538461538463,
      "grad_norm": 24.75,
      "learning_rate": 1.2450000000000001e-05,
      "loss": 0.1293,
      "step": 250
    },
    {
      "epoch": 3.8615384615384616,
      "grad_norm": 26.0,
      "learning_rate": 1.25e-05,
      "loss": 0.2077,
      "step": 251
    },
    {
      "epoch": 3.876923076923077,
      "grad_norm": 21.375,
      "learning_rate": 1.255e-05,
      "loss": 0.1333,
      "step": 252
    },
    {
      "epoch": 3.8923076923076922,
      "grad_norm": 15.8125,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 0.1321,
      "step": 253
    },
    {
      "epoch": 3.9076923076923076,
      "grad_norm": 24.75,
      "learning_rate": 1.2650000000000001e-05,
      "loss": 0.1012,
      "step": 254
    },
    {
      "epoch": 3.9230769230769234,
      "grad_norm": 25.625,
      "learning_rate": 1.27e-05,
      "loss": 0.1758,
      "step": 255
    },
    {
      "epoch": 3.9384615384615387,
      "grad_norm": 18.375,
      "learning_rate": 1.2750000000000002e-05,
      "loss": 0.3081,
      "step": 256
    },
    {
      "epoch": 3.953846153846154,
      "grad_norm": 37.75,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.1647,
      "step": 257
    },
    {
      "epoch": 3.9692307692307693,
      "grad_norm": 44.5,
      "learning_rate": 1.285e-05,
      "loss": 0.1799,
      "step": 258
    },
    {
      "epoch": 3.9846153846153847,
      "grad_norm": 40.75,
      "learning_rate": 1.29e-05,
      "loss": 0.1237,
      "step": 259
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.8125,
      "learning_rate": 1.2950000000000001e-05,
      "loss": 0.1369,
      "step": 260
    },
    {
      "epoch": 4.015384615384615,
      "grad_norm": 21.625,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.1808,
      "step": 261
    },
    {
      "epoch": 4.030769230769231,
      "grad_norm": 28.875,
      "learning_rate": 1.305e-05,
      "loss": 0.2055,
      "step": 262
    },
    {
      "epoch": 4.046153846153846,
      "grad_norm": 26.5,
      "learning_rate": 1.3100000000000002e-05,
      "loss": 0.1013,
      "step": 263
    },
    {
      "epoch": 4.061538461538461,
      "grad_norm": 16.25,
      "learning_rate": 1.3150000000000001e-05,
      "loss": 0.1255,
      "step": 264
    },
    {
      "epoch": 4.076923076923077,
      "grad_norm": 26.125,
      "learning_rate": 1.32e-05,
      "loss": 0.224,
      "step": 265
    },
    {
      "epoch": 4.092307692307692,
      "grad_norm": 12.6875,
      "learning_rate": 1.3250000000000002e-05,
      "loss": 0.1093,
      "step": 266
    },
    {
      "epoch": 4.107692307692307,
      "grad_norm": 47.5,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 0.3022,
      "step": 267
    },
    {
      "epoch": 4.123076923076923,
      "grad_norm": 23.125,
      "learning_rate": 1.3350000000000001e-05,
      "loss": 0.1988,
      "step": 268
    },
    {
      "epoch": 4.138461538461539,
      "grad_norm": 17.75,
      "learning_rate": 1.3400000000000002e-05,
      "loss": 0.0986,
      "step": 269
    },
    {
      "epoch": 4.153846153846154,
      "grad_norm": 9.0,
      "learning_rate": 1.3450000000000002e-05,
      "loss": 0.0828,
      "step": 270
    },
    {
      "epoch": 4.1692307692307695,
      "grad_norm": 19.375,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.1203,
      "step": 271
    },
    {
      "epoch": 4.184615384615385,
      "grad_norm": 15.375,
      "learning_rate": 1.3550000000000002e-05,
      "loss": 0.1274,
      "step": 272
    },
    {
      "epoch": 4.2,
      "grad_norm": 12.5,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.1244,
      "step": 273
    },
    {
      "epoch": 4.2153846153846155,
      "grad_norm": 11.9375,
      "learning_rate": 1.3650000000000001e-05,
      "loss": 0.1542,
      "step": 274
    },
    {
      "epoch": 4.230769230769231,
      "grad_norm": 16.375,
      "learning_rate": 1.3700000000000001e-05,
      "loss": 0.1337,
      "step": 275
    },
    {
      "epoch": 4.246153846153846,
      "grad_norm": 16.875,
      "learning_rate": 1.3750000000000002e-05,
      "loss": 0.2372,
      "step": 276
    },
    {
      "epoch": 4.2615384615384615,
      "grad_norm": 16.75,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 0.1675,
      "step": 277
    },
    {
      "epoch": 4.276923076923077,
      "grad_norm": 17.0,
      "learning_rate": 1.3850000000000001e-05,
      "loss": 0.1235,
      "step": 278
    },
    {
      "epoch": 4.292307692307692,
      "grad_norm": 21.5,
      "learning_rate": 1.3900000000000002e-05,
      "loss": 0.2063,
      "step": 279
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 27.25,
      "learning_rate": 1.3950000000000002e-05,
      "loss": 0.2079,
      "step": 280
    },
    {
      "epoch": 4.323076923076923,
      "grad_norm": 11.3125,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.1538,
      "step": 281
    },
    {
      "epoch": 4.338461538461538,
      "grad_norm": 8.25,
      "learning_rate": 1.4050000000000003e-05,
      "loss": 0.1024,
      "step": 282
    },
    {
      "epoch": 4.3538461538461535,
      "grad_norm": 9.875,
      "learning_rate": 1.4099999999999999e-05,
      "loss": 0.1675,
      "step": 283
    },
    {
      "epoch": 4.36923076923077,
      "grad_norm": 24.0,
      "learning_rate": 1.415e-05,
      "loss": 0.1263,
      "step": 284
    },
    {
      "epoch": 4.384615384615385,
      "grad_norm": 17.875,
      "learning_rate": 1.42e-05,
      "loss": 0.163,
      "step": 285
    },
    {
      "epoch": 4.4,
      "grad_norm": 8.9375,
      "learning_rate": 1.4249999999999999e-05,
      "loss": 0.1083,
      "step": 286
    },
    {
      "epoch": 4.415384615384616,
      "grad_norm": 28.625,
      "learning_rate": 1.43e-05,
      "loss": 0.1998,
      "step": 287
    },
    {
      "epoch": 4.430769230769231,
      "grad_norm": 27.625,
      "learning_rate": 1.435e-05,
      "loss": 0.2076,
      "step": 288
    },
    {
      "epoch": 4.446153846153846,
      "grad_norm": 10.4375,
      "learning_rate": 1.44e-05,
      "loss": 0.1536,
      "step": 289
    },
    {
      "epoch": 4.461538461538462,
      "grad_norm": 20.125,
      "learning_rate": 1.4449999999999999e-05,
      "loss": 0.2528,
      "step": 290
    },
    {
      "epoch": 4.476923076923077,
      "grad_norm": 33.0,
      "learning_rate": 1.45e-05,
      "loss": 0.1661,
      "step": 291
    },
    {
      "epoch": 4.492307692307692,
      "grad_norm": 21.5,
      "learning_rate": 1.455e-05,
      "loss": 0.1965,
      "step": 292
    },
    {
      "epoch": 4.507692307692308,
      "grad_norm": 20.125,
      "learning_rate": 1.4599999999999999e-05,
      "loss": 0.1429,
      "step": 293
    },
    {
      "epoch": 4.523076923076923,
      "grad_norm": 10.75,
      "learning_rate": 1.465e-05,
      "loss": 0.1204,
      "step": 294
    },
    {
      "epoch": 4.538461538461538,
      "grad_norm": 23.25,
      "learning_rate": 1.47e-05,
      "loss": 0.1199,
      "step": 295
    },
    {
      "epoch": 4.553846153846154,
      "grad_norm": 31.75,
      "learning_rate": 1.475e-05,
      "loss": 0.1438,
      "step": 296
    },
    {
      "epoch": 4.569230769230769,
      "grad_norm": 14.3125,
      "learning_rate": 1.48e-05,
      "loss": 0.1534,
      "step": 297
    },
    {
      "epoch": 4.584615384615384,
      "grad_norm": 9.8125,
      "learning_rate": 1.485e-05,
      "loss": 0.179,
      "step": 298
    },
    {
      "epoch": 4.6,
      "grad_norm": 11.0,
      "learning_rate": 1.49e-05,
      "loss": 0.0966,
      "step": 299
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 14.6875,
      "learning_rate": 1.4950000000000001e-05,
      "loss": 0.1236,
      "step": 300
    },
    {
      "epoch": 4.63076923076923,
      "grad_norm": 31.125,
      "learning_rate": 1.5e-05,
      "loss": 0.1927,
      "step": 301
    },
    {
      "epoch": 4.6461538461538465,
      "grad_norm": 21.125,
      "learning_rate": 1.505e-05,
      "loss": 0.1364,
      "step": 302
    },
    {
      "epoch": 4.661538461538462,
      "grad_norm": 18.625,
      "learning_rate": 1.51e-05,
      "loss": 0.1982,
      "step": 303
    },
    {
      "epoch": 4.676923076923077,
      "grad_norm": 35.75,
      "learning_rate": 1.515e-05,
      "loss": 0.2167,
      "step": 304
    },
    {
      "epoch": 4.6923076923076925,
      "grad_norm": 14.8125,
      "learning_rate": 1.52e-05,
      "loss": 0.1212,
      "step": 305
    },
    {
      "epoch": 4.707692307692308,
      "grad_norm": 11.3125,
      "learning_rate": 1.525e-05,
      "loss": 0.0734,
      "step": 306
    },
    {
      "epoch": 4.723076923076923,
      "grad_norm": 11.5,
      "learning_rate": 1.53e-05,
      "loss": 0.0832,
      "step": 307
    },
    {
      "epoch": 4.7384615384615385,
      "grad_norm": 15.6875,
      "learning_rate": 1.535e-05,
      "loss": 0.0666,
      "step": 308
    },
    {
      "epoch": 4.753846153846154,
      "grad_norm": 31.75,
      "learning_rate": 1.54e-05,
      "loss": 0.1434,
      "step": 309
    },
    {
      "epoch": 4.769230769230769,
      "grad_norm": 18.75,
      "learning_rate": 1.545e-05,
      "loss": 0.1699,
      "step": 310
    },
    {
      "epoch": 4.7846153846153845,
      "grad_norm": 28.125,
      "learning_rate": 1.55e-05,
      "loss": 0.2253,
      "step": 311
    },
    {
      "epoch": 4.8,
      "grad_norm": 17.625,
      "learning_rate": 1.5550000000000002e-05,
      "loss": 0.1326,
      "step": 312
    },
    {
      "epoch": 4.815384615384615,
      "grad_norm": 11.625,
      "learning_rate": 1.56e-05,
      "loss": 0.0822,
      "step": 313
    },
    {
      "epoch": 4.8307692307692305,
      "grad_norm": 17.875,
      "learning_rate": 1.565e-05,
      "loss": 0.123,
      "step": 314
    },
    {
      "epoch": 4.846153846153846,
      "grad_norm": 7.625,
      "learning_rate": 1.5700000000000002e-05,
      "loss": 0.1249,
      "step": 315
    },
    {
      "epoch": 4.861538461538462,
      "grad_norm": 7.71875,
      "learning_rate": 1.575e-05,
      "loss": 0.1263,
      "step": 316
    },
    {
      "epoch": 4.876923076923077,
      "grad_norm": 11.0,
      "learning_rate": 1.58e-05,
      "loss": 0.0964,
      "step": 317
    },
    {
      "epoch": 4.892307692307693,
      "grad_norm": 12.625,
      "learning_rate": 1.5850000000000002e-05,
      "loss": 0.0903,
      "step": 318
    },
    {
      "epoch": 4.907692307692308,
      "grad_norm": 5.9375,
      "learning_rate": 1.59e-05,
      "loss": 0.0653,
      "step": 319
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 9.125,
      "learning_rate": 1.595e-05,
      "loss": 0.0987,
      "step": 320
    },
    {
      "epoch": 4.938461538461539,
      "grad_norm": 19.25,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.1145,
      "step": 321
    },
    {
      "epoch": 4.953846153846154,
      "grad_norm": 8.8125,
      "learning_rate": 1.605e-05,
      "loss": 0.081,
      "step": 322
    },
    {
      "epoch": 4.969230769230769,
      "grad_norm": 9.8125,
      "learning_rate": 1.6100000000000002e-05,
      "loss": 0.123,
      "step": 323
    },
    {
      "epoch": 4.984615384615385,
      "grad_norm": 9.25,
      "learning_rate": 1.6150000000000003e-05,
      "loss": 0.1321,
      "step": 324
    },
    {
      "epoch": 5.0,
      "grad_norm": 10.3125,
      "learning_rate": 1.62e-05,
      "loss": 0.1075,
      "step": 325
    },
    {
      "epoch": 5.015384615384615,
      "grad_norm": 11.375,
      "learning_rate": 1.6250000000000002e-05,
      "loss": 0.1406,
      "step": 326
    },
    {
      "epoch": 5.030769230769231,
      "grad_norm": 20.25,
      "learning_rate": 1.63e-05,
      "loss": 0.1537,
      "step": 327
    },
    {
      "epoch": 5.046153846153846,
      "grad_norm": 13.0625,
      "learning_rate": 1.635e-05,
      "loss": 0.0986,
      "step": 328
    },
    {
      "epoch": 5.061538461538461,
      "grad_norm": 18.75,
      "learning_rate": 1.6400000000000002e-05,
      "loss": 0.1683,
      "step": 329
    },
    {
      "epoch": 5.076923076923077,
      "grad_norm": 37.5,
      "learning_rate": 1.645e-05,
      "loss": 0.2011,
      "step": 330
    },
    {
      "epoch": 5.092307692307692,
      "grad_norm": 10.625,
      "learning_rate": 1.65e-05,
      "loss": 0.0945,
      "step": 331
    },
    {
      "epoch": 5.107692307692307,
      "grad_norm": 20.625,
      "learning_rate": 1.6550000000000002e-05,
      "loss": 0.2427,
      "step": 332
    },
    {
      "epoch": 5.123076923076923,
      "grad_norm": 21.75,
      "learning_rate": 1.66e-05,
      "loss": 0.0654,
      "step": 333
    },
    {
      "epoch": 5.138461538461539,
      "grad_norm": 17.625,
      "learning_rate": 1.665e-05,
      "loss": 0.088,
      "step": 334
    },
    {
      "epoch": 5.153846153846154,
      "grad_norm": 8.8125,
      "learning_rate": 1.6700000000000003e-05,
      "loss": 0.0668,
      "step": 335
    },
    {
      "epoch": 5.1692307692307695,
      "grad_norm": 19.0,
      "learning_rate": 1.675e-05,
      "loss": 0.1692,
      "step": 336
    },
    {
      "epoch": 5.184615384615385,
      "grad_norm": 24.5,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.2215,
      "step": 337
    },
    {
      "epoch": 5.2,
      "grad_norm": 28.5,
      "learning_rate": 1.6850000000000003e-05,
      "loss": 0.1764,
      "step": 338
    },
    {
      "epoch": 5.2153846153846155,
      "grad_norm": 17.625,
      "learning_rate": 1.69e-05,
      "loss": 0.1431,
      "step": 339
    },
    {
      "epoch": 5.230769230769231,
      "grad_norm": 11.375,
      "learning_rate": 1.6950000000000002e-05,
      "loss": 0.1128,
      "step": 340
    },
    {
      "epoch": 5.246153846153846,
      "grad_norm": 20.0,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.2252,
      "step": 341
    },
    {
      "epoch": 5.2615384615384615,
      "grad_norm": 8.8125,
      "learning_rate": 1.705e-05,
      "loss": 0.097,
      "step": 342
    },
    {
      "epoch": 5.276923076923077,
      "grad_norm": 26.125,
      "learning_rate": 1.7100000000000002e-05,
      "loss": 0.1297,
      "step": 343
    },
    {
      "epoch": 5.292307692307692,
      "grad_norm": 12.125,
      "learning_rate": 1.7150000000000004e-05,
      "loss": 0.108,
      "step": 344
    },
    {
      "epoch": 5.3076923076923075,
      "grad_norm": 17.625,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 0.1496,
      "step": 345
    },
    {
      "epoch": 5.323076923076923,
      "grad_norm": 10.1875,
      "learning_rate": 1.725e-05,
      "loss": 0.1061,
      "step": 346
    },
    {
      "epoch": 5.338461538461538,
      "grad_norm": 12.4375,
      "learning_rate": 1.73e-05,
      "loss": 0.089,
      "step": 347
    },
    {
      "epoch": 5.3538461538461535,
      "grad_norm": 11.875,
      "learning_rate": 1.7349999999999998e-05,
      "loss": 0.0639,
      "step": 348
    },
    {
      "epoch": 5.36923076923077,
      "grad_norm": 3.96875,
      "learning_rate": 1.74e-05,
      "loss": 0.083,
      "step": 349
    },
    {
      "epoch": 5.384615384615385,
      "grad_norm": 10.125,
      "learning_rate": 1.745e-05,
      "loss": 0.1136,
      "step": 350
    },
    {
      "epoch": 5.4,
      "grad_norm": 11.5,
      "learning_rate": 1.75e-05,
      "loss": 0.1436,
      "step": 351
    },
    {
      "epoch": 5.415384615384616,
      "grad_norm": 14.9375,
      "learning_rate": 1.755e-05,
      "loss": 0.144,
      "step": 352
    },
    {
      "epoch": 5.430769230769231,
      "grad_norm": 6.3125,
      "learning_rate": 1.76e-05,
      "loss": 0.0555,
      "step": 353
    },
    {
      "epoch": 5.446153846153846,
      "grad_norm": 10.9375,
      "learning_rate": 1.765e-05,
      "loss": 0.1173,
      "step": 354
    },
    {
      "epoch": 5.461538461538462,
      "grad_norm": 10.9375,
      "learning_rate": 1.77e-05,
      "loss": 0.1103,
      "step": 355
    },
    {
      "epoch": 5.476923076923077,
      "grad_norm": 11.4375,
      "learning_rate": 1.775e-05,
      "loss": 0.1344,
      "step": 356
    },
    {
      "epoch": 5.492307692307692,
      "grad_norm": 16.125,
      "learning_rate": 1.78e-05,
      "loss": 0.1927,
      "step": 357
    },
    {
      "epoch": 5.507692307692308,
      "grad_norm": 15.25,
      "learning_rate": 1.785e-05,
      "loss": 0.0986,
      "step": 358
    },
    {
      "epoch": 5.523076923076923,
      "grad_norm": 12.1875,
      "learning_rate": 1.79e-05,
      "loss": 0.1172,
      "step": 359
    },
    {
      "epoch": 5.538461538461538,
      "grad_norm": 6.1875,
      "learning_rate": 1.795e-05,
      "loss": 0.1132,
      "step": 360
    },
    {
      "epoch": 5.553846153846154,
      "grad_norm": 10.875,
      "learning_rate": 1.8e-05,
      "loss": 0.1381,
      "step": 361
    },
    {
      "epoch": 5.569230769230769,
      "grad_norm": 8.3125,
      "learning_rate": 1.805e-05,
      "loss": 0.0946,
      "step": 362
    },
    {
      "epoch": 5.584615384615384,
      "grad_norm": 14.375,
      "learning_rate": 1.81e-05,
      "loss": 0.1276,
      "step": 363
    },
    {
      "epoch": 5.6,
      "grad_norm": 10.5,
      "learning_rate": 1.815e-05,
      "loss": 0.0962,
      "step": 364
    },
    {
      "epoch": 5.615384615384615,
      "grad_norm": 8.5625,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 0.0913,
      "step": 365
    },
    {
      "epoch": 5.63076923076923,
      "grad_norm": 7.1875,
      "learning_rate": 1.825e-05,
      "loss": 0.0819,
      "step": 366
    },
    {
      "epoch": 5.6461538461538465,
      "grad_norm": 12.1875,
      "learning_rate": 1.83e-05,
      "loss": 0.1273,
      "step": 367
    },
    {
      "epoch": 5.661538461538462,
      "grad_norm": 9.375,
      "learning_rate": 1.8350000000000002e-05,
      "loss": 0.0837,
      "step": 368
    },
    {
      "epoch": 5.676923076923077,
      "grad_norm": 4.53125,
      "learning_rate": 1.84e-05,
      "loss": 0.0801,
      "step": 369
    },
    {
      "epoch": 5.6923076923076925,
      "grad_norm": 11.3125,
      "learning_rate": 1.845e-05,
      "loss": 0.0621,
      "step": 370
    },
    {
      "epoch": 5.707692307692308,
      "grad_norm": 11.4375,
      "learning_rate": 1.85e-05,
      "loss": 0.1204,
      "step": 371
    },
    {
      "epoch": 5.723076923076923,
      "grad_norm": 11.0625,
      "learning_rate": 1.855e-05,
      "loss": 0.1666,
      "step": 372
    },
    {
      "epoch": 5.7384615384615385,
      "grad_norm": 13.1875,
      "learning_rate": 1.86e-05,
      "loss": 0.1164,
      "step": 373
    },
    {
      "epoch": 5.753846153846154,
      "grad_norm": 7.25,
      "learning_rate": 1.865e-05,
      "loss": 0.1814,
      "step": 374
    },
    {
      "epoch": 5.769230769230769,
      "grad_norm": 8.5625,
      "learning_rate": 1.87e-05,
      "loss": 0.1041,
      "step": 375
    },
    {
      "epoch": 5.7846153846153845,
      "grad_norm": 9.625,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.1143,
      "step": 376
    },
    {
      "epoch": 5.8,
      "grad_norm": 6.59375,
      "learning_rate": 1.88e-05,
      "loss": 0.1254,
      "step": 377
    },
    {
      "epoch": 5.815384615384615,
      "grad_norm": 6.375,
      "learning_rate": 1.885e-05,
      "loss": 0.0716,
      "step": 378
    },
    {
      "epoch": 5.8307692307692305,
      "grad_norm": 6.75,
      "learning_rate": 1.8900000000000002e-05,
      "loss": 0.1265,
      "step": 379
    },
    {
      "epoch": 5.846153846153846,
      "grad_norm": 10.4375,
      "learning_rate": 1.895e-05,
      "loss": 0.1539,
      "step": 380
    },
    {
      "epoch": 5.861538461538462,
      "grad_norm": 10.5625,
      "learning_rate": 1.9e-05,
      "loss": 0.1649,
      "step": 381
    },
    {
      "epoch": 5.876923076923077,
      "grad_norm": 8.625,
      "learning_rate": 1.9050000000000002e-05,
      "loss": 0.0929,
      "step": 382
    },
    {
      "epoch": 5.892307692307693,
      "grad_norm": 6.53125,
      "learning_rate": 1.91e-05,
      "loss": 0.0491,
      "step": 383
    },
    {
      "epoch": 5.907692307692308,
      "grad_norm": 5.75,
      "learning_rate": 1.915e-05,
      "loss": 0.1135,
      "step": 384
    },
    {
      "epoch": 5.923076923076923,
      "grad_norm": 14.6875,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.0993,
      "step": 385
    },
    {
      "epoch": 5.938461538461539,
      "grad_norm": 12.1875,
      "learning_rate": 1.925e-05,
      "loss": 0.1254,
      "step": 386
    },
    {
      "epoch": 5.953846153846154,
      "grad_norm": 12.625,
      "learning_rate": 1.93e-05,
      "loss": 0.084,
      "step": 387
    },
    {
      "epoch": 5.969230769230769,
      "grad_norm": 19.125,
      "learning_rate": 1.9350000000000003e-05,
      "loss": 0.152,
      "step": 388
    },
    {
      "epoch": 5.984615384615385,
      "grad_norm": 7.875,
      "learning_rate": 1.94e-05,
      "loss": 0.1034,
      "step": 389
    },
    {
      "epoch": 6.0,
      "grad_norm": 18.0,
      "learning_rate": 1.9450000000000002e-05,
      "loss": 0.2169,
      "step": 390
    },
    {
      "epoch": 6.015384615384615,
      "grad_norm": 12.625,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 0.0877,
      "step": 391
    },
    {
      "epoch": 6.030769230769231,
      "grad_norm": 4.4375,
      "learning_rate": 1.955e-05,
      "loss": 0.085,
      "step": 392
    },
    {
      "epoch": 6.046153846153846,
      "grad_norm": 5.75,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.0563,
      "step": 393
    },
    {
      "epoch": 6.061538461538461,
      "grad_norm": 8.375,
      "learning_rate": 1.9650000000000003e-05,
      "loss": 0.0846,
      "step": 394
    },
    {
      "epoch": 6.076923076923077,
      "grad_norm": 11.625,
      "learning_rate": 1.97e-05,
      "loss": 0.1454,
      "step": 395
    },
    {
      "epoch": 6.092307692307692,
      "grad_norm": 10.0625,
      "learning_rate": 1.9750000000000002e-05,
      "loss": 0.1314,
      "step": 396
    },
    {
      "epoch": 6.107692307692307,
      "grad_norm": 23.5,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 0.178,
      "step": 397
    },
    {
      "epoch": 6.123076923076923,
      "grad_norm": 10.125,
      "learning_rate": 1.985e-05,
      "loss": 0.118,
      "step": 398
    },
    {
      "epoch": 6.138461538461539,
      "grad_norm": 17.125,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 0.121,
      "step": 399
    },
    {
      "epoch": 6.153846153846154,
      "grad_norm": 14.0625,
      "learning_rate": 1.995e-05,
      "loss": 0.1255,
      "step": 400
    },
    {
      "epoch": 6.1692307692307695,
      "grad_norm": 12.125,
      "learning_rate": 2e-05,
      "loss": 0.1054,
      "step": 401
    },
    {
      "epoch": 6.184615384615385,
      "grad_norm": 9.5625,
      "learning_rate": 2.0050000000000003e-05,
      "loss": 0.1545,
      "step": 402
    },
    {
      "epoch": 6.2,
      "grad_norm": 4.03125,
      "learning_rate": 2.01e-05,
      "loss": 0.126,
      "step": 403
    },
    {
      "epoch": 6.2153846153846155,
      "grad_norm": 14.75,
      "learning_rate": 2.0150000000000002e-05,
      "loss": 0.1196,
      "step": 404
    },
    {
      "epoch": 6.230769230769231,
      "grad_norm": 16.25,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 0.1034,
      "step": 405
    },
    {
      "epoch": 6.246153846153846,
      "grad_norm": 9.4375,
      "learning_rate": 2.025e-05,
      "loss": 0.0715,
      "step": 406
    },
    {
      "epoch": 6.2615384615384615,
      "grad_norm": 11.6875,
      "learning_rate": 2.0300000000000002e-05,
      "loss": 0.0954,
      "step": 407
    },
    {
      "epoch": 6.276923076923077,
      "grad_norm": 5.28125,
      "learning_rate": 2.035e-05,
      "loss": 0.0546,
      "step": 408
    },
    {
      "epoch": 6.292307692307692,
      "grad_norm": 6.34375,
      "learning_rate": 2.04e-05,
      "loss": 0.0659,
      "step": 409
    },
    {
      "epoch": 6.3076923076923075,
      "grad_norm": 4.25,
      "learning_rate": 2.045e-05,
      "loss": 0.0675,
      "step": 410
    },
    {
      "epoch": 6.323076923076923,
      "grad_norm": 20.375,
      "learning_rate": 2.05e-05,
      "loss": 0.1157,
      "step": 411
    },
    {
      "epoch": 6.338461538461538,
      "grad_norm": 9.25,
      "learning_rate": 2.055e-05,
      "loss": 0.0636,
      "step": 412
    },
    {
      "epoch": 6.3538461538461535,
      "grad_norm": 10.1875,
      "learning_rate": 2.06e-05,
      "loss": 0.2038,
      "step": 413
    },
    {
      "epoch": 6.36923076923077,
      "grad_norm": 5.53125,
      "learning_rate": 2.065e-05,
      "loss": 0.0974,
      "step": 414
    },
    {
      "epoch": 6.384615384615385,
      "grad_norm": 4.53125,
      "learning_rate": 2.07e-05,
      "loss": 0.083,
      "step": 415
    },
    {
      "epoch": 6.4,
      "grad_norm": 12.9375,
      "learning_rate": 2.075e-05,
      "loss": 0.1157,
      "step": 416
    },
    {
      "epoch": 6.415384615384616,
      "grad_norm": 7.5,
      "learning_rate": 2.08e-05,
      "loss": 0.078,
      "step": 417
    },
    {
      "epoch": 6.430769230769231,
      "grad_norm": 8.625,
      "learning_rate": 2.085e-05,
      "loss": 0.0725,
      "step": 418
    },
    {
      "epoch": 6.446153846153846,
      "grad_norm": 5.3125,
      "learning_rate": 2.09e-05,
      "loss": 0.0805,
      "step": 419
    },
    {
      "epoch": 6.461538461538462,
      "grad_norm": 14.125,
      "learning_rate": 2.095e-05,
      "loss": 0.0964,
      "step": 420
    },
    {
      "epoch": 6.476923076923077,
      "grad_norm": 4.75,
      "learning_rate": 2.1e-05,
      "loss": 0.0549,
      "step": 421
    },
    {
      "epoch": 6.492307692307692,
      "grad_norm": 11.0,
      "learning_rate": 2.105e-05,
      "loss": 0.1028,
      "step": 422
    },
    {
      "epoch": 6.507692307692308,
      "grad_norm": 15.375,
      "learning_rate": 2.11e-05,
      "loss": 0.0753,
      "step": 423
    },
    {
      "epoch": 6.523076923076923,
      "grad_norm": 22.375,
      "learning_rate": 2.115e-05,
      "loss": 0.1045,
      "step": 424
    },
    {
      "epoch": 6.538461538461538,
      "grad_norm": 7.875,
      "learning_rate": 2.12e-05,
      "loss": 0.1014,
      "step": 425
    },
    {
      "epoch": 6.553846153846154,
      "grad_norm": 9.875,
      "learning_rate": 2.125e-05,
      "loss": 0.1179,
      "step": 426
    },
    {
      "epoch": 6.569230769230769,
      "grad_norm": 4.875,
      "learning_rate": 2.13e-05,
      "loss": 0.0711,
      "step": 427
    },
    {
      "epoch": 6.584615384615384,
      "grad_norm": 5.96875,
      "learning_rate": 2.135e-05,
      "loss": 0.0674,
      "step": 428
    },
    {
      "epoch": 6.6,
      "grad_norm": 20.75,
      "learning_rate": 2.1400000000000002e-05,
      "loss": 0.1309,
      "step": 429
    },
    {
      "epoch": 6.615384615384615,
      "grad_norm": 10.0625,
      "learning_rate": 2.145e-05,
      "loss": 0.0926,
      "step": 430
    },
    {
      "epoch": 6.63076923076923,
      "grad_norm": 6.1875,
      "learning_rate": 2.15e-05,
      "loss": 0.0529,
      "step": 431
    },
    {
      "epoch": 6.6461538461538465,
      "grad_norm": 15.5625,
      "learning_rate": 2.1550000000000002e-05,
      "loss": 0.1532,
      "step": 432
    },
    {
      "epoch": 6.661538461538462,
      "grad_norm": 8.4375,
      "learning_rate": 2.16e-05,
      "loss": 0.0813,
      "step": 433
    },
    {
      "epoch": 6.676923076923077,
      "grad_norm": 5.3125,
      "learning_rate": 2.165e-05,
      "loss": 0.0765,
      "step": 434
    },
    {
      "epoch": 6.6923076923076925,
      "grad_norm": 3.65625,
      "learning_rate": 2.1700000000000002e-05,
      "loss": 0.0728,
      "step": 435
    },
    {
      "epoch": 6.707692307692308,
      "grad_norm": 10.5,
      "learning_rate": 2.175e-05,
      "loss": 0.1672,
      "step": 436
    },
    {
      "epoch": 6.723076923076923,
      "grad_norm": 4.75,
      "learning_rate": 2.18e-05,
      "loss": 0.0662,
      "step": 437
    },
    {
      "epoch": 6.7384615384615385,
      "grad_norm": 14.75,
      "learning_rate": 2.1850000000000003e-05,
      "loss": 0.1928,
      "step": 438
    },
    {
      "epoch": 6.753846153846154,
      "grad_norm": 5.84375,
      "learning_rate": 2.19e-05,
      "loss": 0.159,
      "step": 439
    },
    {
      "epoch": 6.769230769230769,
      "grad_norm": 10.5625,
      "learning_rate": 2.195e-05,
      "loss": 0.1409,
      "step": 440
    },
    {
      "epoch": 6.7846153846153845,
      "grad_norm": 10.625,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.087,
      "step": 441
    },
    {
      "epoch": 6.8,
      "grad_norm": 7.125,
      "learning_rate": 2.205e-05,
      "loss": 0.0745,
      "step": 442
    },
    {
      "epoch": 6.815384615384615,
      "grad_norm": 9.1875,
      "learning_rate": 2.2100000000000002e-05,
      "loss": 0.0826,
      "step": 443
    },
    {
      "epoch": 6.8307692307692305,
      "grad_norm": 4.28125,
      "learning_rate": 2.215e-05,
      "loss": 0.0565,
      "step": 444
    },
    {
      "epoch": 6.846153846153846,
      "grad_norm": 7.34375,
      "learning_rate": 2.22e-05,
      "loss": 0.0849,
      "step": 445
    },
    {
      "epoch": 6.861538461538462,
      "grad_norm": 6.125,
      "learning_rate": 2.2250000000000002e-05,
      "loss": 0.1768,
      "step": 446
    },
    {
      "epoch": 6.876923076923077,
      "grad_norm": 3.8125,
      "learning_rate": 2.23e-05,
      "loss": 0.0806,
      "step": 447
    },
    {
      "epoch": 6.892307692307693,
      "grad_norm": 3.453125,
      "learning_rate": 2.235e-05,
      "loss": 0.1529,
      "step": 448
    },
    {
      "epoch": 6.907692307692308,
      "grad_norm": 6.71875,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 0.0688,
      "step": 449
    },
    {
      "epoch": 6.923076923076923,
      "grad_norm": 4.625,
      "learning_rate": 2.245e-05,
      "loss": 0.0739,
      "step": 450
    },
    {
      "epoch": 6.938461538461539,
      "grad_norm": 4.0625,
      "learning_rate": 2.25e-05,
      "loss": 0.0307,
      "step": 451
    },
    {
      "epoch": 6.953846153846154,
      "grad_norm": 6.96875,
      "learning_rate": 2.2550000000000003e-05,
      "loss": 0.0697,
      "step": 452
    },
    {
      "epoch": 6.969230769230769,
      "grad_norm": 10.25,
      "learning_rate": 2.26e-05,
      "loss": 0.0911,
      "step": 453
    },
    {
      "epoch": 6.984615384615385,
      "grad_norm": 10.8125,
      "learning_rate": 2.265e-05,
      "loss": 0.1453,
      "step": 454
    },
    {
      "epoch": 7.0,
      "grad_norm": 9.0,
      "learning_rate": 2.2700000000000003e-05,
      "loss": 0.099,
      "step": 455
    },
    {
      "epoch": 7.015384615384615,
      "grad_norm": 6.03125,
      "learning_rate": 2.275e-05,
      "loss": 0.0779,
      "step": 456
    },
    {
      "epoch": 7.030769230769231,
      "grad_norm": 2.265625,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 0.0688,
      "step": 457
    },
    {
      "epoch": 7.046153846153846,
      "grad_norm": 7.46875,
      "learning_rate": 2.2850000000000003e-05,
      "loss": 0.077,
      "step": 458
    },
    {
      "epoch": 7.061538461538461,
      "grad_norm": 6.78125,
      "learning_rate": 2.29e-05,
      "loss": 0.1388,
      "step": 459
    },
    {
      "epoch": 7.076923076923077,
      "grad_norm": 4.96875,
      "learning_rate": 2.2950000000000002e-05,
      "loss": 0.0652,
      "step": 460
    },
    {
      "epoch": 7.092307692307692,
      "grad_norm": 3.78125,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.0683,
      "step": 461
    },
    {
      "epoch": 7.107692307692307,
      "grad_norm": 6.15625,
      "learning_rate": 2.305e-05,
      "loss": 0.0734,
      "step": 462
    },
    {
      "epoch": 7.123076923076923,
      "grad_norm": 3.890625,
      "learning_rate": 2.3100000000000002e-05,
      "loss": 0.0547,
      "step": 463
    },
    {
      "epoch": 7.138461538461539,
      "grad_norm": 8.0625,
      "learning_rate": 2.3150000000000004e-05,
      "loss": 0.1036,
      "step": 464
    },
    {
      "epoch": 7.153846153846154,
      "grad_norm": 5.5,
      "learning_rate": 2.32e-05,
      "loss": 0.0996,
      "step": 465
    },
    {
      "epoch": 7.1692307692307695,
      "grad_norm": 5.40625,
      "learning_rate": 2.3250000000000003e-05,
      "loss": 0.0791,
      "step": 466
    },
    {
      "epoch": 7.184615384615385,
      "grad_norm": 7.90625,
      "learning_rate": 2.3300000000000004e-05,
      "loss": 0.093,
      "step": 467
    },
    {
      "epoch": 7.2,
      "grad_norm": 1.7265625,
      "learning_rate": 2.3350000000000002e-05,
      "loss": 0.0309,
      "step": 468
    },
    {
      "epoch": 7.2153846153846155,
      "grad_norm": 2.0625,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 0.0848,
      "step": 469
    },
    {
      "epoch": 7.230769230769231,
      "grad_norm": 14.1875,
      "learning_rate": 2.345e-05,
      "loss": 0.091,
      "step": 470
    },
    {
      "epoch": 7.246153846153846,
      "grad_norm": 5.71875,
      "learning_rate": 2.35e-05,
      "loss": 0.0727,
      "step": 471
    },
    {
      "epoch": 7.2615384615384615,
      "grad_norm": 4.96875,
      "learning_rate": 2.355e-05,
      "loss": 0.0331,
      "step": 472
    },
    {
      "epoch": 7.276923076923077,
      "grad_norm": 3.609375,
      "learning_rate": 2.36e-05,
      "loss": 0.0924,
      "step": 473
    },
    {
      "epoch": 7.292307692307692,
      "grad_norm": 4.75,
      "learning_rate": 2.365e-05,
      "loss": 0.1412,
      "step": 474
    },
    {
      "epoch": 7.3076923076923075,
      "grad_norm": 13.0625,
      "learning_rate": 2.37e-05,
      "loss": 0.0921,
      "step": 475
    },
    {
      "epoch": 7.323076923076923,
      "grad_norm": 5.0,
      "learning_rate": 2.375e-05,
      "loss": 0.0612,
      "step": 476
    },
    {
      "epoch": 7.338461538461538,
      "grad_norm": 11.375,
      "learning_rate": 2.38e-05,
      "loss": 0.105,
      "step": 477
    },
    {
      "epoch": 7.3538461538461535,
      "grad_norm": 2.875,
      "learning_rate": 2.385e-05,
      "loss": 0.0433,
      "step": 478
    },
    {
      "epoch": 7.36923076923077,
      "grad_norm": 6.53125,
      "learning_rate": 2.39e-05,
      "loss": 0.0196,
      "step": 479
    },
    {
      "epoch": 7.384615384615385,
      "grad_norm": 11.125,
      "learning_rate": 2.395e-05,
      "loss": 0.093,
      "step": 480
    },
    {
      "epoch": 7.4,
      "grad_norm": 5.34375,
      "learning_rate": 2.4e-05,
      "loss": 0.1338,
      "step": 481
    },
    {
      "epoch": 7.415384615384616,
      "grad_norm": 12.0625,
      "learning_rate": 2.4050000000000002e-05,
      "loss": 0.072,
      "step": 482
    },
    {
      "epoch": 7.430769230769231,
      "grad_norm": 6.6875,
      "learning_rate": 2.41e-05,
      "loss": 0.1559,
      "step": 483
    },
    {
      "epoch": 7.446153846153846,
      "grad_norm": 7.96875,
      "learning_rate": 2.415e-05,
      "loss": 0.0772,
      "step": 484
    },
    {
      "epoch": 7.461538461538462,
      "grad_norm": 6.6875,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 0.0992,
      "step": 485
    },
    {
      "epoch": 7.476923076923077,
      "grad_norm": 11.875,
      "learning_rate": 2.425e-05,
      "loss": 0.0804,
      "step": 486
    },
    {
      "epoch": 7.492307692307692,
      "grad_norm": 18.875,
      "learning_rate": 2.43e-05,
      "loss": 0.0928,
      "step": 487
    },
    {
      "epoch": 7.507692307692308,
      "grad_norm": 3.640625,
      "learning_rate": 2.435e-05,
      "loss": 0.103,
      "step": 488
    },
    {
      "epoch": 7.523076923076923,
      "grad_norm": 13.3125,
      "learning_rate": 2.44e-05,
      "loss": 0.1347,
      "step": 489
    },
    {
      "epoch": 7.538461538461538,
      "grad_norm": 3.4375,
      "learning_rate": 2.445e-05,
      "loss": 0.0601,
      "step": 490
    },
    {
      "epoch": 7.553846153846154,
      "grad_norm": 3.015625,
      "learning_rate": 2.45e-05,
      "loss": 0.0377,
      "step": 491
    },
    {
      "epoch": 7.569230769230769,
      "grad_norm": 9.375,
      "learning_rate": 2.455e-05,
      "loss": 0.139,
      "step": 492
    },
    {
      "epoch": 7.584615384615384,
      "grad_norm": 5.9375,
      "learning_rate": 2.46e-05,
      "loss": 0.0481,
      "step": 493
    },
    {
      "epoch": 7.6,
      "grad_norm": 14.375,
      "learning_rate": 2.465e-05,
      "loss": 0.0804,
      "step": 494
    },
    {
      "epoch": 7.615384615384615,
      "grad_norm": 4.59375,
      "learning_rate": 2.47e-05,
      "loss": 0.0367,
      "step": 495
    },
    {
      "epoch": 7.63076923076923,
      "grad_norm": 12.3125,
      "learning_rate": 2.4750000000000002e-05,
      "loss": 0.1189,
      "step": 496
    },
    {
      "epoch": 7.6461538461538465,
      "grad_norm": 3.640625,
      "learning_rate": 2.48e-05,
      "loss": 0.1133,
      "step": 497
    },
    {
      "epoch": 7.661538461538462,
      "grad_norm": 7.53125,
      "learning_rate": 2.485e-05,
      "loss": 0.07,
      "step": 498
    },
    {
      "epoch": 7.676923076923077,
      "grad_norm": 5.15625,
      "learning_rate": 2.4900000000000002e-05,
      "loss": 0.0562,
      "step": 499
    },
    {
      "epoch": 7.6923076923076925,
      "grad_norm": 4.78125,
      "learning_rate": 2.495e-05,
      "loss": 0.0915,
      "step": 500
    },
    {
      "epoch": 7.707692307692308,
      "grad_norm": 1.890625,
      "learning_rate": 2.5e-05,
      "loss": 0.0612,
      "step": 501
    },
    {
      "epoch": 7.723076923076923,
      "grad_norm": 3.546875,
      "learning_rate": 2.4999996953825926e-05,
      "loss": 0.0414,
      "step": 502
    },
    {
      "epoch": 7.7384615384615385,
      "grad_norm": 3.859375,
      "learning_rate": 2.499998781530519e-05,
      "loss": 0.0437,
      "step": 503
    },
    {
      "epoch": 7.753846153846154,
      "grad_norm": 11.375,
      "learning_rate": 2.4999972584442245e-05,
      "loss": 0.1625,
      "step": 504
    },
    {
      "epoch": 7.769230769230769,
      "grad_norm": 10.375,
      "learning_rate": 2.4999951261244508e-05,
      "loss": 0.0529,
      "step": 505
    },
    {
      "epoch": 7.7846153846153845,
      "grad_norm": 6.84375,
      "learning_rate": 2.499992384572238e-05,
      "loss": 0.0428,
      "step": 506
    },
    {
      "epoch": 7.8,
      "grad_norm": 17.375,
      "learning_rate": 2.499989033788922e-05,
      "loss": 0.1582,
      "step": 507
    },
    {
      "epoch": 7.815384615384615,
      "grad_norm": 3.015625,
      "learning_rate": 2.499985073776136e-05,
      "loss": 0.0306,
      "step": 508
    },
    {
      "epoch": 7.8307692307692305,
      "grad_norm": 6.90625,
      "learning_rate": 2.4999805045358104e-05,
      "loss": 0.0773,
      "step": 509
    },
    {
      "epoch": 7.846153846153846,
      "grad_norm": 5.4375,
      "learning_rate": 2.4999753260701714e-05,
      "loss": 0.0966,
      "step": 510
    },
    {
      "epoch": 7.861538461538462,
      "grad_norm": 5.46875,
      "learning_rate": 2.4999695383817435e-05,
      "loss": 0.0665,
      "step": 511
    },
    {
      "epoch": 7.876923076923077,
      "grad_norm": 7.40625,
      "learning_rate": 2.4999631414733476e-05,
      "loss": 0.0786,
      "step": 512
    },
    {
      "epoch": 7.892307692307693,
      "grad_norm": 4.3125,
      "learning_rate": 2.499956135348101e-05,
      "loss": 0.0932,
      "step": 513
    },
    {
      "epoch": 7.907692307692308,
      "grad_norm": 5.9375,
      "learning_rate": 2.4999485200094186e-05,
      "loss": 0.0484,
      "step": 514
    },
    {
      "epoch": 7.923076923076923,
      "grad_norm": 4.03125,
      "learning_rate": 2.4999402954610125e-05,
      "loss": 0.0812,
      "step": 515
    },
    {
      "epoch": 7.938461538461539,
      "grad_norm": 4.125,
      "learning_rate": 2.4999314617068904e-05,
      "loss": 0.022,
      "step": 516
    },
    {
      "epoch": 7.953846153846154,
      "grad_norm": 1.9375,
      "learning_rate": 2.4999220187513583e-05,
      "loss": 0.0262,
      "step": 517
    },
    {
      "epoch": 7.969230769230769,
      "grad_norm": 11.5625,
      "learning_rate": 2.4999119665990183e-05,
      "loss": 0.0179,
      "step": 518
    },
    {
      "epoch": 7.984615384615385,
      "grad_norm": 12.375,
      "learning_rate": 2.4999013052547703e-05,
      "loss": 0.0714,
      "step": 519
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.34375,
      "learning_rate": 2.4998900347238096e-05,
      "loss": 0.0518,
      "step": 520
    },
    {
      "epoch": 8.015384615384615,
      "grad_norm": 6.96875,
      "learning_rate": 2.4998781550116305e-05,
      "loss": 0.1216,
      "step": 521
    },
    {
      "epoch": 8.03076923076923,
      "grad_norm": 7.09375,
      "learning_rate": 2.499865666124022e-05,
      "loss": 0.0711,
      "step": 522
    },
    {
      "epoch": 8.046153846153846,
      "grad_norm": 10.25,
      "learning_rate": 2.4998525680670712e-05,
      "loss": 0.1403,
      "step": 523
    },
    {
      "epoch": 8.061538461538461,
      "grad_norm": 8.5,
      "learning_rate": 2.4998388608471625e-05,
      "loss": 0.0794,
      "step": 524
    },
    {
      "epoch": 8.076923076923077,
      "grad_norm": 9.8125,
      "learning_rate": 2.4998245444709757e-05,
      "loss": 0.1028,
      "step": 525
    },
    {
      "epoch": 8.092307692307692,
      "grad_norm": 6.59375,
      "learning_rate": 2.4998096189454893e-05,
      "loss": 0.1788,
      "step": 526
    },
    {
      "epoch": 8.107692307692307,
      "grad_norm": 6.34375,
      "learning_rate": 2.499794084277977e-05,
      "loss": 0.0874,
      "step": 527
    },
    {
      "epoch": 8.123076923076923,
      "grad_norm": 4.09375,
      "learning_rate": 2.4997779404760114e-05,
      "loss": 0.0638,
      "step": 528
    },
    {
      "epoch": 8.138461538461538,
      "grad_norm": 3.109375,
      "learning_rate": 2.4997611875474595e-05,
      "loss": 0.0544,
      "step": 529
    },
    {
      "epoch": 8.153846153846153,
      "grad_norm": 5.875,
      "learning_rate": 2.4997438255004876e-05,
      "loss": 0.1179,
      "step": 530
    },
    {
      "epoch": 8.169230769230769,
      "grad_norm": 2.5,
      "learning_rate": 2.499725854343557e-05,
      "loss": 0.0732,
      "step": 531
    },
    {
      "epoch": 8.184615384615384,
      "grad_norm": 7.15625,
      "learning_rate": 2.4997072740854265e-05,
      "loss": 0.0635,
      "step": 532
    },
    {
      "epoch": 8.2,
      "grad_norm": 17.625,
      "learning_rate": 2.499688084735153e-05,
      "loss": 0.0971,
      "step": 533
    },
    {
      "epoch": 8.215384615384615,
      "grad_norm": 6.21875,
      "learning_rate": 2.4996682863020875e-05,
      "loss": 0.1969,
      "step": 534
    },
    {
      "epoch": 8.23076923076923,
      "grad_norm": 8.1875,
      "learning_rate": 2.499647878795881e-05,
      "loss": 0.1148,
      "step": 535
    },
    {
      "epoch": 8.246153846153845,
      "grad_norm": 6.65625,
      "learning_rate": 2.499626862226479e-05,
      "loss": 0.0959,
      "step": 536
    },
    {
      "epoch": 8.261538461538462,
      "grad_norm": 5.125,
      "learning_rate": 2.499605236604125e-05,
      "loss": 0.1067,
      "step": 537
    },
    {
      "epoch": 8.276923076923078,
      "grad_norm": 12.125,
      "learning_rate": 2.499583001939359e-05,
      "loss": 0.1205,
      "step": 538
    },
    {
      "epoch": 8.292307692307693,
      "grad_norm": 3.078125,
      "learning_rate": 2.4995601582430183e-05,
      "loss": 0.0862,
      "step": 539
    },
    {
      "epoch": 8.307692307692308,
      "grad_norm": 5.875,
      "learning_rate": 2.499536705526236e-05,
      "loss": 0.0989,
      "step": 540
    },
    {
      "epoch": 8.323076923076924,
      "grad_norm": 2.515625,
      "learning_rate": 2.499512643800443e-05,
      "loss": 0.0834,
      "step": 541
    },
    {
      "epoch": 8.338461538461539,
      "grad_norm": 3.125,
      "learning_rate": 2.499487973077367e-05,
      "loss": 0.0609,
      "step": 542
    },
    {
      "epoch": 8.353846153846154,
      "grad_norm": 5.15625,
      "learning_rate": 2.4994626933690314e-05,
      "loss": 0.0672,
      "step": 543
    },
    {
      "epoch": 8.36923076923077,
      "grad_norm": 6.03125,
      "learning_rate": 2.499436804687758e-05,
      "loss": 0.1548,
      "step": 544
    },
    {
      "epoch": 8.384615384615385,
      "grad_norm": 2.203125,
      "learning_rate": 2.499410307046164e-05,
      "loss": 0.0408,
      "step": 545
    },
    {
      "epoch": 8.4,
      "grad_norm": 2.84375,
      "learning_rate": 2.4993832004571646e-05,
      "loss": 0.0588,
      "step": 546
    },
    {
      "epoch": 8.415384615384616,
      "grad_norm": 7.75,
      "learning_rate": 2.499355484933971e-05,
      "loss": 0.085,
      "step": 547
    },
    {
      "epoch": 8.430769230769231,
      "grad_norm": 4.21875,
      "learning_rate": 2.4993271604900913e-05,
      "loss": 0.0753,
      "step": 548
    },
    {
      "epoch": 8.446153846153846,
      "grad_norm": 0.8203125,
      "learning_rate": 2.4992982271393307e-05,
      "loss": 0.0585,
      "step": 549
    },
    {
      "epoch": 8.461538461538462,
      "grad_norm": 15.625,
      "learning_rate": 2.4992686848957905e-05,
      "loss": 0.0637,
      "step": 550
    },
    {
      "epoch": 8.476923076923077,
      "grad_norm": 7.28125,
      "learning_rate": 2.49923853377387e-05,
      "loss": 0.1055,
      "step": 551
    },
    {
      "epoch": 8.492307692307692,
      "grad_norm": 5.59375,
      "learning_rate": 2.4992077737882636e-05,
      "loss": 0.08,
      "step": 552
    },
    {
      "epoch": 8.507692307692308,
      "grad_norm": 8.4375,
      "learning_rate": 2.499176404953964e-05,
      "loss": 0.1613,
      "step": 553
    },
    {
      "epoch": 8.523076923076923,
      "grad_norm": 11.4375,
      "learning_rate": 2.49914442728626e-05,
      "loss": 0.091,
      "step": 554
    },
    {
      "epoch": 8.538461538461538,
      "grad_norm": 19.125,
      "learning_rate": 2.4991118408007366e-05,
      "loss": 0.113,
      "step": 555
    },
    {
      "epoch": 8.553846153846154,
      "grad_norm": 4.8125,
      "learning_rate": 2.4990786455132764e-05,
      "loss": 0.0763,
      "step": 556
    },
    {
      "epoch": 8.569230769230769,
      "grad_norm": 2.90625,
      "learning_rate": 2.4990448414400585e-05,
      "loss": 0.0346,
      "step": 557
    },
    {
      "epoch": 8.584615384615384,
      "grad_norm": 5.625,
      "learning_rate": 2.4990104285975586e-05,
      "loss": 0.0585,
      "step": 558
    },
    {
      "epoch": 8.6,
      "grad_norm": 10.125,
      "learning_rate": 2.498975407002549e-05,
      "loss": 0.0587,
      "step": 559
    },
    {
      "epoch": 8.615384615384615,
      "grad_norm": 7.21875,
      "learning_rate": 2.4989397766720983e-05,
      "loss": 0.0742,
      "step": 560
    },
    {
      "epoch": 8.63076923076923,
      "grad_norm": 2.0,
      "learning_rate": 2.498903537623573e-05,
      "loss": 0.0489,
      "step": 561
    },
    {
      "epoch": 8.646153846153846,
      "grad_norm": 7.84375,
      "learning_rate": 2.4988666898746356e-05,
      "loss": 0.1261,
      "step": 562
    },
    {
      "epoch": 8.661538461538461,
      "grad_norm": 5.0625,
      "learning_rate": 2.4988292334432444e-05,
      "loss": 0.0502,
      "step": 563
    },
    {
      "epoch": 8.676923076923076,
      "grad_norm": 6.03125,
      "learning_rate": 2.4987911683476562e-05,
      "loss": 0.0922,
      "step": 564
    },
    {
      "epoch": 8.692307692307692,
      "grad_norm": 6.46875,
      "learning_rate": 2.4987524946064228e-05,
      "loss": 0.08,
      "step": 565
    },
    {
      "epoch": 8.707692307692307,
      "grad_norm": 1.6640625,
      "learning_rate": 2.4987132122383936e-05,
      "loss": 0.0224,
      "step": 566
    },
    {
      "epoch": 8.723076923076922,
      "grad_norm": 4.40625,
      "learning_rate": 2.4986733212627143e-05,
      "loss": 0.1247,
      "step": 567
    },
    {
      "epoch": 8.73846153846154,
      "grad_norm": 1.5625,
      "learning_rate": 2.498632821698827e-05,
      "loss": 0.117,
      "step": 568
    },
    {
      "epoch": 8.753846153846155,
      "grad_norm": 4.28125,
      "learning_rate": 2.4985917135664714e-05,
      "loss": 0.0432,
      "step": 569
    },
    {
      "epoch": 8.76923076923077,
      "grad_norm": 6.96875,
      "learning_rate": 2.4985499968856824e-05,
      "loss": 0.0564,
      "step": 570
    },
    {
      "epoch": 8.784615384615385,
      "grad_norm": 6.5,
      "learning_rate": 2.4985076716767927e-05,
      "loss": 0.1109,
      "step": 571
    },
    {
      "epoch": 8.8,
      "grad_norm": 6.71875,
      "learning_rate": 2.498464737960431e-05,
      "loss": 0.062,
      "step": 572
    },
    {
      "epoch": 8.815384615384616,
      "grad_norm": 8.0625,
      "learning_rate": 2.498421195757522e-05,
      "loss": 0.0723,
      "step": 573
    },
    {
      "epoch": 8.830769230769231,
      "grad_norm": 6.84375,
      "learning_rate": 2.4983770450892884e-05,
      "loss": 0.0999,
      "step": 574
    },
    {
      "epoch": 8.846153846153847,
      "grad_norm": 4.5,
      "learning_rate": 2.4983322859772484e-05,
      "loss": 0.0564,
      "step": 575
    },
    {
      "epoch": 8.861538461538462,
      "grad_norm": 2.921875,
      "learning_rate": 2.4982869184432174e-05,
      "loss": 0.0638,
      "step": 576
    },
    {
      "epoch": 8.876923076923077,
      "grad_norm": 7.78125,
      "learning_rate": 2.4982409425093067e-05,
      "loss": 0.056,
      "step": 577
    },
    {
      "epoch": 8.892307692307693,
      "grad_norm": 4.65625,
      "learning_rate": 2.498194358197924e-05,
      "loss": 0.1124,
      "step": 578
    },
    {
      "epoch": 8.907692307692308,
      "grad_norm": 2.984375,
      "learning_rate": 2.498147165531775e-05,
      "loss": 0.1051,
      "step": 579
    },
    {
      "epoch": 8.923076923076923,
      "grad_norm": 3.78125,
      "learning_rate": 2.4980993645338595e-05,
      "loss": 0.0459,
      "step": 580
    },
    {
      "epoch": 8.938461538461539,
      "grad_norm": 12.9375,
      "learning_rate": 2.4980509552274765e-05,
      "loss": 0.0798,
      "step": 581
    },
    {
      "epoch": 8.953846153846154,
      "grad_norm": 10.5,
      "learning_rate": 2.498001937636219e-05,
      "loss": 0.1361,
      "step": 582
    },
    {
      "epoch": 8.96923076923077,
      "grad_norm": 5.625,
      "learning_rate": 2.4979523117839783e-05,
      "loss": 0.0582,
      "step": 583
    },
    {
      "epoch": 8.984615384615385,
      "grad_norm": 5.5625,
      "learning_rate": 2.4979020776949405e-05,
      "loss": 0.1189,
      "step": 584
    },
    {
      "epoch": 9.0,
      "grad_norm": 4.78125,
      "learning_rate": 2.4978512353935903e-05,
      "loss": 0.055,
      "step": 585
    },
    {
      "epoch": 9.015384615384615,
      "grad_norm": 3.296875,
      "learning_rate": 2.497799784904707e-05,
      "loss": 0.0545,
      "step": 586
    },
    {
      "epoch": 9.03076923076923,
      "grad_norm": 7.3125,
      "learning_rate": 2.4977477262533667e-05,
      "loss": 0.0702,
      "step": 587
    },
    {
      "epoch": 9.046153846153846,
      "grad_norm": 2.65625,
      "learning_rate": 2.4976950594649423e-05,
      "loss": 0.0374,
      "step": 588
    },
    {
      "epoch": 9.061538461538461,
      "grad_norm": 4.03125,
      "learning_rate": 2.4976417845651033e-05,
      "loss": 0.0543,
      "step": 589
    },
    {
      "epoch": 9.076923076923077,
      "grad_norm": 3.671875,
      "learning_rate": 2.4975879015798152e-05,
      "loss": 0.0513,
      "step": 590
    },
    {
      "epoch": 9.092307692307692,
      "grad_norm": 2.671875,
      "learning_rate": 2.4975334105353396e-05,
      "loss": 0.036,
      "step": 591
    },
    {
      "epoch": 9.107692307692307,
      "grad_norm": 4.96875,
      "learning_rate": 2.4974783114582352e-05,
      "loss": 0.0399,
      "step": 592
    },
    {
      "epoch": 9.123076923076923,
      "grad_norm": 3.765625,
      "learning_rate": 2.4974226043753558e-05,
      "loss": 0.0354,
      "step": 593
    },
    {
      "epoch": 9.138461538461538,
      "grad_norm": 4.03125,
      "learning_rate": 2.497366289313853e-05,
      "loss": 0.1398,
      "step": 594
    },
    {
      "epoch": 9.153846153846153,
      "grad_norm": 3.265625,
      "learning_rate": 2.4973093663011743e-05,
      "loss": 0.0322,
      "step": 595
    },
    {
      "epoch": 9.169230769230769,
      "grad_norm": 7.15625,
      "learning_rate": 2.4972518353650626e-05,
      "loss": 0.0536,
      "step": 596
    },
    {
      "epoch": 9.184615384615384,
      "grad_norm": 5.34375,
      "learning_rate": 2.4971936965335585e-05,
      "loss": 0.0424,
      "step": 597
    },
    {
      "epoch": 9.2,
      "grad_norm": 4.375,
      "learning_rate": 2.497134949834997e-05,
      "loss": 0.0709,
      "step": 598
    },
    {
      "epoch": 9.215384615384615,
      "grad_norm": 13.25,
      "learning_rate": 2.4970755952980125e-05,
      "loss": 0.0442,
      "step": 599
    },
    {
      "epoch": 9.23076923076923,
      "grad_norm": 7.0,
      "learning_rate": 2.4970156329515315e-05,
      "loss": 0.0295,
      "step": 600
    },
    {
      "epoch": 9.246153846153845,
      "grad_norm": 6.71875,
      "learning_rate": 2.4969550628247805e-05,
      "loss": 0.0589,
      "step": 601
    },
    {
      "epoch": 9.261538461538462,
      "grad_norm": 8.5,
      "learning_rate": 2.4968938849472797e-05,
      "loss": 0.055,
      "step": 602
    },
    {
      "epoch": 9.276923076923078,
      "grad_norm": 2.34375,
      "learning_rate": 2.496832099348847e-05,
      "loss": 0.0699,
      "step": 603
    },
    {
      "epoch": 9.292307692307693,
      "grad_norm": 5.28125,
      "learning_rate": 2.4967697060595958e-05,
      "loss": 0.058,
      "step": 604
    },
    {
      "epoch": 9.307692307692308,
      "grad_norm": 6.3125,
      "learning_rate": 2.4967067051099356e-05,
      "loss": 0.0984,
      "step": 605
    },
    {
      "epoch": 9.323076923076924,
      "grad_norm": 2.484375,
      "learning_rate": 2.4966430965305727e-05,
      "loss": 0.0804,
      "step": 606
    },
    {
      "epoch": 9.338461538461539,
      "grad_norm": 1.53125,
      "learning_rate": 2.4965788803525088e-05,
      "loss": 0.0294,
      "step": 607
    },
    {
      "epoch": 9.353846153846154,
      "grad_norm": 2.90625,
      "learning_rate": 2.4965140566070423e-05,
      "loss": 0.0562,
      "step": 608
    },
    {
      "epoch": 9.36923076923077,
      "grad_norm": 9.875,
      "learning_rate": 2.4964486253257674e-05,
      "loss": 0.0348,
      "step": 609
    },
    {
      "epoch": 9.384615384615385,
      "grad_norm": 11.875,
      "learning_rate": 2.4963825865405747e-05,
      "loss": 0.0792,
      "step": 610
    },
    {
      "epoch": 9.4,
      "grad_norm": 2.359375,
      "learning_rate": 2.4963159402836506e-05,
      "loss": 0.0807,
      "step": 611
    },
    {
      "epoch": 9.415384615384616,
      "grad_norm": 7.4375,
      "learning_rate": 2.4962486865874774e-05,
      "loss": 0.116,
      "step": 612
    },
    {
      "epoch": 9.430769230769231,
      "grad_norm": 4.53125,
      "learning_rate": 2.4961808254848342e-05,
      "loss": 0.0511,
      "step": 613
    },
    {
      "epoch": 9.446153846153846,
      "grad_norm": 6.125,
      "learning_rate": 2.496112357008795e-05,
      "loss": 0.0682,
      "step": 614
    },
    {
      "epoch": 9.461538461538462,
      "grad_norm": 6.21875,
      "learning_rate": 2.4960432811927308e-05,
      "loss": 0.0649,
      "step": 615
    },
    {
      "epoch": 9.476923076923077,
      "grad_norm": 12.9375,
      "learning_rate": 2.495973598070309e-05,
      "loss": 0.0838,
      "step": 616
    },
    {
      "epoch": 9.492307692307692,
      "grad_norm": 7.1875,
      "learning_rate": 2.4959033076754917e-05,
      "loss": 0.0671,
      "step": 617
    },
    {
      "epoch": 9.507692307692308,
      "grad_norm": 11.5,
      "learning_rate": 2.495832410042537e-05,
      "loss": 0.0734,
      "step": 618
    },
    {
      "epoch": 9.523076923076923,
      "grad_norm": 6.8125,
      "learning_rate": 2.495760905206001e-05,
      "loss": 0.0726,
      "step": 619
    },
    {
      "epoch": 9.538461538461538,
      "grad_norm": 7.375,
      "learning_rate": 2.4956887932007332e-05,
      "loss": 0.0425,
      "step": 620
    },
    {
      "epoch": 9.553846153846154,
      "grad_norm": 3.859375,
      "learning_rate": 2.4956160740618806e-05,
      "loss": 0.0438,
      "step": 621
    },
    {
      "epoch": 9.569230769230769,
      "grad_norm": 4.78125,
      "learning_rate": 2.4955427478248848e-05,
      "loss": 0.0967,
      "step": 622
    },
    {
      "epoch": 9.584615384615384,
      "grad_norm": 2.546875,
      "learning_rate": 2.4954688145254855e-05,
      "loss": 0.0254,
      "step": 623
    },
    {
      "epoch": 9.6,
      "grad_norm": 3.796875,
      "learning_rate": 2.495394274199716e-05,
      "loss": 0.0652,
      "step": 624
    },
    {
      "epoch": 9.615384615384615,
      "grad_norm": 3.65625,
      "learning_rate": 2.495319126883906e-05,
      "loss": 0.0411,
      "step": 625
    },
    {
      "epoch": 9.63076923076923,
      "grad_norm": 5.75,
      "learning_rate": 2.495243372614682e-05,
      "loss": 0.0657,
      "step": 626
    },
    {
      "epoch": 9.646153846153846,
      "grad_norm": 8.4375,
      "learning_rate": 2.495167011428966e-05,
      "loss": 0.0665,
      "step": 627
    },
    {
      "epoch": 9.661538461538461,
      "grad_norm": 3.78125,
      "learning_rate": 2.4950900433639746e-05,
      "loss": 0.0607,
      "step": 628
    },
    {
      "epoch": 9.676923076923076,
      "grad_norm": 2.390625,
      "learning_rate": 2.4950124684572217e-05,
      "loss": 0.0426,
      "step": 629
    },
    {
      "epoch": 9.692307692307692,
      "grad_norm": 5.09375,
      "learning_rate": 2.494934286746516e-05,
      "loss": 0.0701,
      "step": 630
    },
    {
      "epoch": 9.707692307692307,
      "grad_norm": 8.5625,
      "learning_rate": 2.494855498269963e-05,
      "loss": 0.114,
      "step": 631
    },
    {
      "epoch": 9.723076923076922,
      "grad_norm": 7.46875,
      "learning_rate": 2.4947761030659627e-05,
      "loss": 0.0712,
      "step": 632
    },
    {
      "epoch": 9.73846153846154,
      "grad_norm": 2.875,
      "learning_rate": 2.4946961011732118e-05,
      "loss": 0.0218,
      "step": 633
    },
    {
      "epoch": 9.753846153846155,
      "grad_norm": 5.3125,
      "learning_rate": 2.4946154926307014e-05,
      "loss": 0.0693,
      "step": 634
    },
    {
      "epoch": 9.76923076923077,
      "grad_norm": 4.34375,
      "learning_rate": 2.4945342774777196e-05,
      "loss": 0.0519,
      "step": 635
    },
    {
      "epoch": 9.784615384615385,
      "grad_norm": 5.90625,
      "learning_rate": 2.4944524557538503e-05,
      "loss": 0.0828,
      "step": 636
    },
    {
      "epoch": 9.8,
      "grad_norm": 3.90625,
      "learning_rate": 2.4943700274989714e-05,
      "loss": 0.0546,
      "step": 637
    },
    {
      "epoch": 9.815384615384616,
      "grad_norm": 4.21875,
      "learning_rate": 2.494286992753258e-05,
      "loss": 0.1179,
      "step": 638
    },
    {
      "epoch": 9.830769230769231,
      "grad_norm": 1.0703125,
      "learning_rate": 2.49420335155718e-05,
      "loss": 0.0431,
      "step": 639
    },
    {
      "epoch": 9.846153846153847,
      "grad_norm": 2.0625,
      "learning_rate": 2.4941191039515032e-05,
      "loss": 0.0257,
      "step": 640
    },
    {
      "epoch": 9.861538461538462,
      "grad_norm": 1.3359375,
      "learning_rate": 2.494034249977289e-05,
      "loss": 0.0417,
      "step": 641
    },
    {
      "epoch": 9.876923076923077,
      "grad_norm": 6.46875,
      "learning_rate": 2.493948789675894e-05,
      "loss": 0.0604,
      "step": 642
    },
    {
      "epoch": 9.892307692307693,
      "grad_norm": 4.5,
      "learning_rate": 2.4938627230889705e-05,
      "loss": 0.0298,
      "step": 643
    },
    {
      "epoch": 9.907692307692308,
      "grad_norm": 2.8125,
      "learning_rate": 2.4937760502584666e-05,
      "loss": 0.0362,
      "step": 644
    },
    {
      "epoch": 9.923076923076923,
      "grad_norm": 3.9375,
      "learning_rate": 2.4936887712266254e-05,
      "loss": 0.0401,
      "step": 645
    },
    {
      "epoch": 9.938461538461539,
      "grad_norm": 4.0,
      "learning_rate": 2.4936008860359854e-05,
      "loss": 0.1089,
      "step": 646
    },
    {
      "epoch": 9.953846153846154,
      "grad_norm": 4.78125,
      "learning_rate": 2.493512394729381e-05,
      "loss": 0.0513,
      "step": 647
    },
    {
      "epoch": 9.96923076923077,
      "grad_norm": 5.5625,
      "learning_rate": 2.4934232973499415e-05,
      "loss": 0.0537,
      "step": 648
    },
    {
      "epoch": 9.984615384615385,
      "grad_norm": 5.125,
      "learning_rate": 2.493333593941093e-05,
      "loss": 0.041,
      "step": 649
    },
    {
      "epoch": 10.0,
      "grad_norm": 4.53125,
      "learning_rate": 2.4932432845465543e-05,
      "loss": 0.0426,
      "step": 650
    },
    {
      "epoch": 10.015384615384615,
      "grad_norm": 5.40625,
      "learning_rate": 2.4931523692103418e-05,
      "loss": 0.0632,
      "step": 651
    },
    {
      "epoch": 10.03076923076923,
      "grad_norm": 7.84375,
      "learning_rate": 2.4930608479767667e-05,
      "loss": 0.0394,
      "step": 652
    },
    {
      "epoch": 10.046153846153846,
      "grad_norm": 3.3125,
      "learning_rate": 2.492968720890435e-05,
      "loss": 0.0365,
      "step": 653
    },
    {
      "epoch": 10.061538461538461,
      "grad_norm": 0.9921875,
      "learning_rate": 2.4928759879962487e-05,
      "loss": 0.0246,
      "step": 654
    },
    {
      "epoch": 10.076923076923077,
      "grad_norm": 9.0625,
      "learning_rate": 2.4927826493394046e-05,
      "loss": 0.039,
      "step": 655
    },
    {
      "epoch": 10.092307692307692,
      "grad_norm": 3.375,
      "learning_rate": 2.4926887049653943e-05,
      "loss": 0.0266,
      "step": 656
    },
    {
      "epoch": 10.107692307692307,
      "grad_norm": 1.5078125,
      "learning_rate": 2.492594154920006e-05,
      "loss": 0.0215,
      "step": 657
    },
    {
      "epoch": 10.123076923076923,
      "grad_norm": 2.0,
      "learning_rate": 2.4924989992493213e-05,
      "loss": 0.0508,
      "step": 658
    },
    {
      "epoch": 10.138461538461538,
      "grad_norm": 6.21875,
      "learning_rate": 2.4924032379997186e-05,
      "loss": 0.0631,
      "step": 659
    },
    {
      "epoch": 10.153846153846153,
      "grad_norm": 10.375,
      "learning_rate": 2.492306871217871e-05,
      "loss": 0.1035,
      "step": 660
    },
    {
      "epoch": 10.169230769230769,
      "grad_norm": 1.953125,
      "learning_rate": 2.4922098989507454e-05,
      "loss": 0.0224,
      "step": 661
    },
    {
      "epoch": 10.184615384615384,
      "grad_norm": 3.609375,
      "learning_rate": 2.492112321245606e-05,
      "loss": 0.048,
      "step": 662
    },
    {
      "epoch": 10.2,
      "grad_norm": 2.234375,
      "learning_rate": 2.4920141381500107e-05,
      "loss": 0.0647,
      "step": 663
    },
    {
      "epoch": 10.215384615384615,
      "grad_norm": 4.8125,
      "learning_rate": 2.4919153497118126e-05,
      "loss": 0.0346,
      "step": 664
    },
    {
      "epoch": 10.23076923076923,
      "grad_norm": 1.9609375,
      "learning_rate": 2.4918159559791596e-05,
      "loss": 0.0248,
      "step": 665
    },
    {
      "epoch": 10.246153846153845,
      "grad_norm": 0.64453125,
      "learning_rate": 2.4917159570004954e-05,
      "loss": 0.0183,
      "step": 666
    },
    {
      "epoch": 10.261538461538462,
      "grad_norm": 4.71875,
      "learning_rate": 2.491615352824559e-05,
      "loss": 0.0657,
      "step": 667
    },
    {
      "epoch": 10.276923076923078,
      "grad_norm": 7.96875,
      "learning_rate": 2.4915141435003823e-05,
      "loss": 0.0839,
      "step": 668
    },
    {
      "epoch": 10.292307692307693,
      "grad_norm": 5.71875,
      "learning_rate": 2.4914123290772945e-05,
      "loss": 0.06,
      "step": 669
    },
    {
      "epoch": 10.307692307692308,
      "grad_norm": 2.671875,
      "learning_rate": 2.491309909604918e-05,
      "loss": 0.0347,
      "step": 670
    },
    {
      "epoch": 10.323076923076924,
      "grad_norm": 3.90625,
      "learning_rate": 2.491206885133171e-05,
      "loss": 0.083,
      "step": 671
    },
    {
      "epoch": 10.338461538461539,
      "grad_norm": 2.90625,
      "learning_rate": 2.491103255712267e-05,
      "loss": 0.0518,
      "step": 672
    },
    {
      "epoch": 10.353846153846154,
      "grad_norm": 4.125,
      "learning_rate": 2.4909990213927135e-05,
      "loss": 0.0641,
      "step": 673
    },
    {
      "epoch": 10.36923076923077,
      "grad_norm": 7.75,
      "learning_rate": 2.4908941822253124e-05,
      "loss": 0.0505,
      "step": 674
    },
    {
      "epoch": 10.384615384615385,
      "grad_norm": 2.0,
      "learning_rate": 2.4907887382611613e-05,
      "loss": 0.021,
      "step": 675
    },
    {
      "epoch": 10.4,
      "grad_norm": 1.9375,
      "learning_rate": 2.4906826895516528e-05,
      "loss": 0.0373,
      "step": 676
    },
    {
      "epoch": 10.415384615384616,
      "grad_norm": 2.40625,
      "learning_rate": 2.4905760361484733e-05,
      "loss": 0.1129,
      "step": 677
    },
    {
      "epoch": 10.430769230769231,
      "grad_norm": 2.46875,
      "learning_rate": 2.490468778103604e-05,
      "loss": 0.0321,
      "step": 678
    },
    {
      "epoch": 10.446153846153846,
      "grad_norm": 6.03125,
      "learning_rate": 2.4903609154693222e-05,
      "loss": 0.1234,
      "step": 679
    },
    {
      "epoch": 10.461538461538462,
      "grad_norm": 4.28125,
      "learning_rate": 2.490252448298198e-05,
      "loss": 0.0497,
      "step": 680
    },
    {
      "epoch": 10.476923076923077,
      "grad_norm": 1.734375,
      "learning_rate": 2.4901433766430975e-05,
      "loss": 0.0167,
      "step": 681
    },
    {
      "epoch": 10.492307692307692,
      "grad_norm": 3.8125,
      "learning_rate": 2.4900337005571807e-05,
      "loss": 0.0364,
      "step": 682
    },
    {
      "epoch": 10.507692307692308,
      "grad_norm": 4.375,
      "learning_rate": 2.489923420093902e-05,
      "loss": 0.0665,
      "step": 683
    },
    {
      "epoch": 10.523076923076923,
      "grad_norm": 3.15625,
      "learning_rate": 2.4898125353070112e-05,
      "loss": 0.0949,
      "step": 684
    },
    {
      "epoch": 10.538461538461538,
      "grad_norm": 1.2578125,
      "learning_rate": 2.4897010462505522e-05,
      "loss": 0.0342,
      "step": 685
    },
    {
      "epoch": 10.553846153846154,
      "grad_norm": 5.78125,
      "learning_rate": 2.489588952978863e-05,
      "loss": 0.0334,
      "step": 686
    },
    {
      "epoch": 10.569230769230769,
      "grad_norm": 2.75,
      "learning_rate": 2.4894762555465772e-05,
      "loss": 0.0561,
      "step": 687
    },
    {
      "epoch": 10.584615384615384,
      "grad_norm": 2.609375,
      "learning_rate": 2.4893629540086216e-05,
      "loss": 0.0321,
      "step": 688
    },
    {
      "epoch": 10.6,
      "grad_norm": 2.046875,
      "learning_rate": 2.489249048420218e-05,
      "loss": 0.0661,
      "step": 689
    },
    {
      "epoch": 10.615384615384615,
      "grad_norm": 5.09375,
      "learning_rate": 2.4891345388368834e-05,
      "loss": 0.0659,
      "step": 690
    },
    {
      "epoch": 10.63076923076923,
      "grad_norm": 2.984375,
      "learning_rate": 2.489019425314427e-05,
      "loss": 0.0345,
      "step": 691
    },
    {
      "epoch": 10.646153846153846,
      "grad_norm": 4.8125,
      "learning_rate": 2.488903707908955e-05,
      "loss": 0.0536,
      "step": 692
    },
    {
      "epoch": 10.661538461538461,
      "grad_norm": 0.87890625,
      "learning_rate": 2.488787386676866e-05,
      "loss": 0.0158,
      "step": 693
    },
    {
      "epoch": 10.676923076923076,
      "grad_norm": 0.78125,
      "learning_rate": 2.4886704616748536e-05,
      "loss": 0.0265,
      "step": 694
    },
    {
      "epoch": 10.692307692307692,
      "grad_norm": 10.9375,
      "learning_rate": 2.488552932959906e-05,
      "loss": 0.0977,
      "step": 695
    },
    {
      "epoch": 10.707692307692307,
      "grad_norm": 11.75,
      "learning_rate": 2.4884348005893045e-05,
      "loss": 0.0882,
      "step": 696
    },
    {
      "epoch": 10.723076923076922,
      "grad_norm": 6.0,
      "learning_rate": 2.488316064620626e-05,
      "loss": 0.0282,
      "step": 697
    },
    {
      "epoch": 10.73846153846154,
      "grad_norm": 5.15625,
      "learning_rate": 2.488196725111741e-05,
      "loss": 0.0585,
      "step": 698
    },
    {
      "epoch": 10.753846153846155,
      "grad_norm": 7.25,
      "learning_rate": 2.488076782120814e-05,
      "loss": 0.0956,
      "step": 699
    },
    {
      "epoch": 10.76923076923077,
      "grad_norm": 5.875,
      "learning_rate": 2.4879562357063037e-05,
      "loss": 0.0467,
      "step": 700
    },
    {
      "epoch": 10.784615384615385,
      "grad_norm": 2.9375,
      "learning_rate": 2.487835085926963e-05,
      "loss": 0.0464,
      "step": 701
    },
    {
      "epoch": 10.8,
      "grad_norm": 5.15625,
      "learning_rate": 2.487713332841839e-05,
      "loss": 0.0451,
      "step": 702
    },
    {
      "epoch": 10.815384615384616,
      "grad_norm": 5.90625,
      "learning_rate": 2.487590976510272e-05,
      "loss": 0.0382,
      "step": 703
    },
    {
      "epoch": 10.830769230769231,
      "grad_norm": 6.4375,
      "learning_rate": 2.4874680169918976e-05,
      "loss": 0.0609,
      "step": 704
    },
    {
      "epoch": 10.846153846153847,
      "grad_norm": 1.28125,
      "learning_rate": 2.4873444543466448e-05,
      "loss": 0.0328,
      "step": 705
    },
    {
      "epoch": 10.861538461538462,
      "grad_norm": 12.4375,
      "learning_rate": 2.4872202886347362e-05,
      "loss": 0.1514,
      "step": 706
    },
    {
      "epoch": 10.876923076923077,
      "grad_norm": 5.65625,
      "learning_rate": 2.487095519916689e-05,
      "loss": 0.0934,
      "step": 707
    },
    {
      "epoch": 10.892307692307693,
      "grad_norm": 2.28125,
      "learning_rate": 2.4869701482533138e-05,
      "loss": 0.0357,
      "step": 708
    },
    {
      "epoch": 10.907692307692308,
      "grad_norm": 3.8125,
      "learning_rate": 2.4868441737057153e-05,
      "loss": 0.0383,
      "step": 709
    },
    {
      "epoch": 10.923076923076923,
      "grad_norm": 4.71875,
      "learning_rate": 2.486717596335292e-05,
      "loss": 0.0472,
      "step": 710
    },
    {
      "epoch": 10.938461538461539,
      "grad_norm": 5.78125,
      "learning_rate": 2.4865904162037358e-05,
      "loss": 0.0326,
      "step": 711
    },
    {
      "epoch": 10.953846153846154,
      "grad_norm": 3.703125,
      "learning_rate": 2.4864626333730328e-05,
      "loss": 0.0422,
      "step": 712
    },
    {
      "epoch": 10.96923076923077,
      "grad_norm": 5.25,
      "learning_rate": 2.4863342479054633e-05,
      "loss": 0.0838,
      "step": 713
    },
    {
      "epoch": 10.984615384615385,
      "grad_norm": 8.0,
      "learning_rate": 2.4862052598636005e-05,
      "loss": 0.1194,
      "step": 714
    },
    {
      "epoch": 11.0,
      "grad_norm": 6.40625,
      "learning_rate": 2.4860756693103114e-05,
      "loss": 0.0519,
      "step": 715
    },
    {
      "epoch": 11.015384615384615,
      "grad_norm": 4.84375,
      "learning_rate": 2.4859454763087577e-05,
      "loss": 0.0398,
      "step": 716
    },
    {
      "epoch": 11.03076923076923,
      "grad_norm": 5.96875,
      "learning_rate": 2.4858146809223925e-05,
      "loss": 0.0904,
      "step": 717
    },
    {
      "epoch": 11.046153846153846,
      "grad_norm": 5.25,
      "learning_rate": 2.485683283214965e-05,
      "loss": 0.0958,
      "step": 718
    },
    {
      "epoch": 11.061538461538461,
      "grad_norm": 2.984375,
      "learning_rate": 2.485551283250517e-05,
      "loss": 0.016,
      "step": 719
    },
    {
      "epoch": 11.076923076923077,
      "grad_norm": 1.8125,
      "learning_rate": 2.4854186810933828e-05,
      "loss": 0.0141,
      "step": 720
    },
    {
      "epoch": 11.092307692307692,
      "grad_norm": 10.3125,
      "learning_rate": 2.4852854768081912e-05,
      "loss": 0.0651,
      "step": 721
    },
    {
      "epoch": 11.107692307692307,
      "grad_norm": 11.1875,
      "learning_rate": 2.485151670459865e-05,
      "loss": 0.0782,
      "step": 722
    },
    {
      "epoch": 11.123076923076923,
      "grad_norm": 7.71875,
      "learning_rate": 2.485017262113619e-05,
      "loss": 0.0708,
      "step": 723
    },
    {
      "epoch": 11.138461538461538,
      "grad_norm": 5.875,
      "learning_rate": 2.484882251834963e-05,
      "loss": 0.0604,
      "step": 724
    },
    {
      "epoch": 11.153846153846153,
      "grad_norm": 7.78125,
      "learning_rate": 2.484746639689699e-05,
      "loss": 0.0717,
      "step": 725
    },
    {
      "epoch": 11.169230769230769,
      "grad_norm": 6.78125,
      "learning_rate": 2.4846104257439222e-05,
      "loss": 0.0859,
      "step": 726
    },
    {
      "epoch": 11.184615384615384,
      "grad_norm": 6.53125,
      "learning_rate": 2.4844736100640224e-05,
      "loss": 0.0626,
      "step": 727
    },
    {
      "epoch": 11.2,
      "grad_norm": 3.8125,
      "learning_rate": 2.484336192716682e-05,
      "loss": 0.0415,
      "step": 728
    },
    {
      "epoch": 11.215384615384615,
      "grad_norm": 5.3125,
      "learning_rate": 2.4841981737688754e-05,
      "loss": 0.0449,
      "step": 729
    },
    {
      "epoch": 11.23076923076923,
      "grad_norm": 3.671875,
      "learning_rate": 2.4840595532878727e-05,
      "loss": 0.0424,
      "step": 730
    },
    {
      "epoch": 11.246153846153845,
      "grad_norm": 5.90625,
      "learning_rate": 2.483920331341235e-05,
      "loss": 0.0741,
      "step": 731
    },
    {
      "epoch": 11.261538461538462,
      "grad_norm": 3.34375,
      "learning_rate": 2.4837805079968176e-05,
      "loss": 0.0825,
      "step": 732
    },
    {
      "epoch": 11.276923076923078,
      "grad_norm": 3.34375,
      "learning_rate": 2.483640083322769e-05,
      "loss": 0.0447,
      "step": 733
    },
    {
      "epoch": 11.292307692307693,
      "grad_norm": 4.53125,
      "learning_rate": 2.48349905738753e-05,
      "loss": 0.0923,
      "step": 734
    },
    {
      "epoch": 11.307692307692308,
      "grad_norm": 3.203125,
      "learning_rate": 2.483357430259835e-05,
      "loss": 0.0598,
      "step": 735
    },
    {
      "epoch": 11.323076923076924,
      "grad_norm": 6.65625,
      "learning_rate": 2.483215202008712e-05,
      "loss": 0.1078,
      "step": 736
    },
    {
      "epoch": 11.338461538461539,
      "grad_norm": 3.984375,
      "learning_rate": 2.4830723727034805e-05,
      "loss": 0.0544,
      "step": 737
    },
    {
      "epoch": 11.353846153846154,
      "grad_norm": 5.0,
      "learning_rate": 2.4829289424137538e-05,
      "loss": 0.0625,
      "step": 738
    },
    {
      "epoch": 11.36923076923077,
      "grad_norm": 1.09375,
      "learning_rate": 2.4827849112094386e-05,
      "loss": 0.009,
      "step": 739
    },
    {
      "epoch": 11.384615384615385,
      "grad_norm": 7.40625,
      "learning_rate": 2.482640279160734e-05,
      "loss": 0.0638,
      "step": 740
    },
    {
      "epoch": 11.4,
      "grad_norm": 5.53125,
      "learning_rate": 2.4824950463381314e-05,
      "loss": 0.103,
      "step": 741
    },
    {
      "epoch": 11.415384615384616,
      "grad_norm": 5.5,
      "learning_rate": 2.4823492128124154e-05,
      "loss": 0.0774,
      "step": 742
    },
    {
      "epoch": 11.430769230769231,
      "grad_norm": 3.6875,
      "learning_rate": 2.482202778654664e-05,
      "loss": 0.0675,
      "step": 743
    },
    {
      "epoch": 11.446153846153846,
      "grad_norm": 4.03125,
      "learning_rate": 2.4820557439362478e-05,
      "loss": 0.0372,
      "step": 744
    },
    {
      "epoch": 11.461538461538462,
      "grad_norm": 1.875,
      "learning_rate": 2.4819081087288286e-05,
      "loss": 0.0353,
      "step": 745
    },
    {
      "epoch": 11.476923076923077,
      "grad_norm": 0.6015625,
      "learning_rate": 2.481759873104363e-05,
      "loss": 0.0545,
      "step": 746
    },
    {
      "epoch": 11.492307692307692,
      "grad_norm": 1.984375,
      "learning_rate": 2.4816110371350982e-05,
      "loss": 0.0628,
      "step": 747
    },
    {
      "epoch": 11.507692307692308,
      "grad_norm": 2.9375,
      "learning_rate": 2.4814616008935762e-05,
      "loss": 0.0424,
      "step": 748
    },
    {
      "epoch": 11.523076923076923,
      "grad_norm": 8.9375,
      "learning_rate": 2.4813115644526298e-05,
      "loss": 0.1104,
      "step": 749
    },
    {
      "epoch": 11.538461538461538,
      "grad_norm": 5.5625,
      "learning_rate": 2.4811609278853847e-05,
      "loss": 0.0369,
      "step": 750
    },
    {
      "epoch": 11.553846153846154,
      "grad_norm": 3.140625,
      "learning_rate": 2.4810096912652604e-05,
      "loss": 0.0516,
      "step": 751
    },
    {
      "epoch": 11.569230769230769,
      "grad_norm": 3.90625,
      "learning_rate": 2.4808578546659663e-05,
      "loss": 0.0293,
      "step": 752
    },
    {
      "epoch": 11.584615384615384,
      "grad_norm": 6.4375,
      "learning_rate": 2.4807054181615068e-05,
      "loss": 0.0828,
      "step": 753
    },
    {
      "epoch": 11.6,
      "grad_norm": 2.78125,
      "learning_rate": 2.4805523818261774e-05,
      "loss": 0.063,
      "step": 754
    },
    {
      "epoch": 11.615384615384615,
      "grad_norm": 6.65625,
      "learning_rate": 2.4803987457345658e-05,
      "loss": 0.0642,
      "step": 755
    },
    {
      "epoch": 11.63076923076923,
      "grad_norm": 12.1875,
      "learning_rate": 2.4802445099615525e-05,
      "loss": 0.0785,
      "step": 756
    },
    {
      "epoch": 11.646153846153846,
      "grad_norm": 3.953125,
      "learning_rate": 2.4800896745823105e-05,
      "loss": 0.0402,
      "step": 757
    },
    {
      "epoch": 11.661538461538461,
      "grad_norm": 5.59375,
      "learning_rate": 2.4799342396723043e-05,
      "loss": 0.1111,
      "step": 758
    },
    {
      "epoch": 11.676923076923076,
      "grad_norm": 2.21875,
      "learning_rate": 2.479778205307291e-05,
      "loss": 0.0523,
      "step": 759
    },
    {
      "epoch": 11.692307692307692,
      "grad_norm": 4.5,
      "learning_rate": 2.4796215715633196e-05,
      "loss": 0.0606,
      "step": 760
    },
    {
      "epoch": 11.707692307692307,
      "grad_norm": 3.21875,
      "learning_rate": 2.4794643385167327e-05,
      "loss": 0.0498,
      "step": 761
    },
    {
      "epoch": 11.723076923076922,
      "grad_norm": 2.234375,
      "learning_rate": 2.479306506244162e-05,
      "loss": 0.0765,
      "step": 762
    },
    {
      "epoch": 11.73846153846154,
      "grad_norm": 2.609375,
      "learning_rate": 2.4791480748225344e-05,
      "loss": 0.0373,
      "step": 763
    },
    {
      "epoch": 11.753846153846155,
      "grad_norm": 8.4375,
      "learning_rate": 2.478989044329067e-05,
      "loss": 0.0774,
      "step": 764
    },
    {
      "epoch": 11.76923076923077,
      "grad_norm": 3.59375,
      "learning_rate": 2.4788294148412693e-05,
      "loss": 0.0252,
      "step": 765
    },
    {
      "epoch": 11.784615384615385,
      "grad_norm": 4.1875,
      "learning_rate": 2.478669186436943e-05,
      "loss": 0.0535,
      "step": 766
    },
    {
      "epoch": 11.8,
      "grad_norm": 2.921875,
      "learning_rate": 2.478508359194181e-05,
      "loss": 0.0652,
      "step": 767
    },
    {
      "epoch": 11.815384615384616,
      "grad_norm": 3.34375,
      "learning_rate": 2.4783469331913688e-05,
      "loss": 0.0368,
      "step": 768
    },
    {
      "epoch": 11.830769230769231,
      "grad_norm": 2.328125,
      "learning_rate": 2.4781849085071833e-05,
      "loss": 0.0283,
      "step": 769
    },
    {
      "epoch": 11.846153846153847,
      "grad_norm": 1.4453125,
      "learning_rate": 2.478022285220594e-05,
      "loss": 0.0583,
      "step": 770
    },
    {
      "epoch": 11.861538461538462,
      "grad_norm": 5.03125,
      "learning_rate": 2.4778590634108613e-05,
      "loss": 0.0386,
      "step": 771
    },
    {
      "epoch": 11.876923076923077,
      "grad_norm": 2.40625,
      "learning_rate": 2.4776952431575364e-05,
      "loss": 0.0252,
      "step": 772
    },
    {
      "epoch": 11.892307692307693,
      "grad_norm": 3.140625,
      "learning_rate": 2.477530824540465e-05,
      "loss": 0.0436,
      "step": 773
    },
    {
      "epoch": 11.907692307692308,
      "grad_norm": 3.765625,
      "learning_rate": 2.4773658076397816e-05,
      "loss": 0.0341,
      "step": 774
    },
    {
      "epoch": 11.923076923076923,
      "grad_norm": 4.09375,
      "learning_rate": 2.4772001925359133e-05,
      "loss": 0.0513,
      "step": 775
    },
    {
      "epoch": 11.938461538461539,
      "grad_norm": 2.375,
      "learning_rate": 2.4770339793095802e-05,
      "loss": 0.0192,
      "step": 776
    },
    {
      "epoch": 11.953846153846154,
      "grad_norm": 3.03125,
      "learning_rate": 2.4768671680417914e-05,
      "loss": 0.0534,
      "step": 777
    },
    {
      "epoch": 11.96923076923077,
      "grad_norm": 3.9375,
      "learning_rate": 2.476699758813849e-05,
      "loss": 0.0234,
      "step": 778
    },
    {
      "epoch": 11.984615384615385,
      "grad_norm": 2.859375,
      "learning_rate": 2.476531751707346e-05,
      "loss": 0.034,
      "step": 779
    },
    {
      "epoch": 12.0,
      "grad_norm": 7.90625,
      "learning_rate": 2.4763631468041675e-05,
      "loss": 0.0412,
      "step": 780
    },
    {
      "epoch": 12.015384615384615,
      "grad_norm": 3.578125,
      "learning_rate": 2.4761939441864895e-05,
      "loss": 0.0211,
      "step": 781
    },
    {
      "epoch": 12.03076923076923,
      "grad_norm": 4.1875,
      "learning_rate": 2.476024143936779e-05,
      "loss": 0.031,
      "step": 782
    },
    {
      "epoch": 12.046153846153846,
      "grad_norm": 0.71875,
      "learning_rate": 2.4758537461377947e-05,
      "loss": 0.0097,
      "step": 783
    },
    {
      "epoch": 12.061538461538461,
      "grad_norm": 14.8125,
      "learning_rate": 2.475682750872586e-05,
      "loss": 0.0357,
      "step": 784
    },
    {
      "epoch": 12.076923076923077,
      "grad_norm": 4.78125,
      "learning_rate": 2.475511158224495e-05,
      "loss": 0.0291,
      "step": 785
    },
    {
      "epoch": 12.092307692307692,
      "grad_norm": 3.703125,
      "learning_rate": 2.4753389682771523e-05,
      "loss": 0.0471,
      "step": 786
    },
    {
      "epoch": 12.107692307692307,
      "grad_norm": 3.390625,
      "learning_rate": 2.4751661811144828e-05,
      "loss": 0.0287,
      "step": 787
    },
    {
      "epoch": 12.123076923076923,
      "grad_norm": 8.9375,
      "learning_rate": 2.4749927968206996e-05,
      "loss": 0.0871,
      "step": 788
    },
    {
      "epoch": 12.138461538461538,
      "grad_norm": 2.109375,
      "learning_rate": 2.4748188154803086e-05,
      "loss": 0.0597,
      "step": 789
    },
    {
      "epoch": 12.153846153846153,
      "grad_norm": 0.94140625,
      "learning_rate": 2.4746442371781066e-05,
      "loss": 0.0291,
      "step": 790
    },
    {
      "epoch": 12.169230769230769,
      "grad_norm": 2.5625,
      "learning_rate": 2.47446906199918e-05,
      "loss": 0.0287,
      "step": 791
    },
    {
      "epoch": 12.184615384615384,
      "grad_norm": 9.375,
      "learning_rate": 2.474293290028908e-05,
      "loss": 0.0594,
      "step": 792
    },
    {
      "epoch": 12.2,
      "grad_norm": 1.984375,
      "learning_rate": 2.4741169213529594e-05,
      "loss": 0.0834,
      "step": 793
    },
    {
      "epoch": 12.215384615384615,
      "grad_norm": 4.15625,
      "learning_rate": 2.473939956057294e-05,
      "loss": 0.0506,
      "step": 794
    },
    {
      "epoch": 12.23076923076923,
      "grad_norm": 5.1875,
      "learning_rate": 2.4737623942281624e-05,
      "loss": 0.0222,
      "step": 795
    },
    {
      "epoch": 12.246153846153845,
      "grad_norm": 6.71875,
      "learning_rate": 2.4735842359521064e-05,
      "loss": 0.0438,
      "step": 796
    },
    {
      "epoch": 12.261538461538462,
      "grad_norm": 4.6875,
      "learning_rate": 2.473405481315958e-05,
      "loss": 0.0871,
      "step": 797
    },
    {
      "epoch": 12.276923076923078,
      "grad_norm": 6.34375,
      "learning_rate": 2.4732261304068403e-05,
      "loss": 0.0614,
      "step": 798
    },
    {
      "epoch": 12.292307692307693,
      "grad_norm": 2.734375,
      "learning_rate": 2.4730461833121664e-05,
      "loss": 0.0452,
      "step": 799
    },
    {
      "epoch": 12.307692307692308,
      "grad_norm": 2.15625,
      "learning_rate": 2.4728656401196406e-05,
      "loss": 0.0174,
      "step": 800
    },
    {
      "epoch": 12.323076923076924,
      "grad_norm": 4.4375,
      "learning_rate": 2.4726845009172572e-05,
      "loss": 0.05,
      "step": 801
    },
    {
      "epoch": 12.338461538461539,
      "grad_norm": 2.859375,
      "learning_rate": 2.4725027657933013e-05,
      "loss": 0.0144,
      "step": 802
    },
    {
      "epoch": 12.353846153846154,
      "grad_norm": 4.75,
      "learning_rate": 2.4723204348363487e-05,
      "loss": 0.0328,
      "step": 803
    },
    {
      "epoch": 12.36923076923077,
      "grad_norm": 3.421875,
      "learning_rate": 2.4721375081352648e-05,
      "loss": 0.0294,
      "step": 804
    },
    {
      "epoch": 12.384615384615385,
      "grad_norm": 6.03125,
      "learning_rate": 2.4719539857792068e-05,
      "loss": 0.0491,
      "step": 805
    },
    {
      "epoch": 12.4,
      "grad_norm": 0.5546875,
      "learning_rate": 2.47176986785762e-05,
      "loss": 0.0395,
      "step": 806
    },
    {
      "epoch": 12.415384615384616,
      "grad_norm": 1.203125,
      "learning_rate": 2.471585154460242e-05,
      "loss": 0.0461,
      "step": 807
    },
    {
      "epoch": 12.430769230769231,
      "grad_norm": 3.703125,
      "learning_rate": 2.4713998456771e-05,
      "loss": 0.0874,
      "step": 808
    },
    {
      "epoch": 12.446153846153846,
      "grad_norm": 5.59375,
      "learning_rate": 2.4712139415985104e-05,
      "loss": 0.0643,
      "step": 809
    },
    {
      "epoch": 12.461538461538462,
      "grad_norm": 4.71875,
      "learning_rate": 2.471027442315081e-05,
      "loss": 0.0539,
      "step": 810
    },
    {
      "epoch": 12.476923076923077,
      "grad_norm": 11.5,
      "learning_rate": 2.47084034791771e-05,
      "loss": 0.1044,
      "step": 811
    },
    {
      "epoch": 12.492307692307692,
      "grad_norm": 9.5625,
      "learning_rate": 2.4706526584975843e-05,
      "loss": 0.0814,
      "step": 812
    },
    {
      "epoch": 12.507692307692308,
      "grad_norm": 3.984375,
      "learning_rate": 2.4704643741461815e-05,
      "loss": 0.0211,
      "step": 813
    },
    {
      "epoch": 12.523076923076923,
      "grad_norm": 1.84375,
      "learning_rate": 2.4702754949552686e-05,
      "loss": 0.032,
      "step": 814
    },
    {
      "epoch": 12.538461538461538,
      "grad_norm": 3.15625,
      "learning_rate": 2.470086021016904e-05,
      "loss": 0.0321,
      "step": 815
    },
    {
      "epoch": 12.553846153846154,
      "grad_norm": 3.109375,
      "learning_rate": 2.4698959524234346e-05,
      "loss": 0.0639,
      "step": 816
    },
    {
      "epoch": 12.569230769230769,
      "grad_norm": 3.546875,
      "learning_rate": 2.4697052892674972e-05,
      "loss": 0.0801,
      "step": 817
    },
    {
      "epoch": 12.584615384615384,
      "grad_norm": 2.453125,
      "learning_rate": 2.469514031642019e-05,
      "loss": 0.0443,
      "step": 818
    },
    {
      "epoch": 12.6,
      "grad_norm": 3.984375,
      "learning_rate": 2.4693221796402166e-05,
      "loss": 0.062,
      "step": 819
    },
    {
      "epoch": 12.615384615384615,
      "grad_norm": 5.21875,
      "learning_rate": 2.4691297333555964e-05,
      "loss": 0.0901,
      "step": 820
    },
    {
      "epoch": 12.63076923076923,
      "grad_norm": 2.109375,
      "learning_rate": 2.468936692881954e-05,
      "loss": 0.0416,
      "step": 821
    },
    {
      "epoch": 12.646153846153846,
      "grad_norm": 3.296875,
      "learning_rate": 2.468743058313376e-05,
      "loss": 0.094,
      "step": 822
    },
    {
      "epoch": 12.661538461538461,
      "grad_norm": 5.53125,
      "learning_rate": 2.4685488297442362e-05,
      "loss": 0.0816,
      "step": 823
    },
    {
      "epoch": 12.676923076923076,
      "grad_norm": 2.296875,
      "learning_rate": 2.4683540072692003e-05,
      "loss": 0.0481,
      "step": 824
    },
    {
      "epoch": 12.692307692307692,
      "grad_norm": 3.359375,
      "learning_rate": 2.4681585909832217e-05,
      "loss": 0.0555,
      "step": 825
    },
    {
      "epoch": 12.707692307692307,
      "grad_norm": 3.375,
      "learning_rate": 2.4679625809815443e-05,
      "loss": 0.0503,
      "step": 826
    },
    {
      "epoch": 12.723076923076922,
      "grad_norm": 5.90625,
      "learning_rate": 2.4677659773597007e-05,
      "loss": 0.067,
      "step": 827
    },
    {
      "epoch": 12.73846153846154,
      "grad_norm": 4.5625,
      "learning_rate": 2.4675687802135135e-05,
      "loss": 0.0345,
      "step": 828
    },
    {
      "epoch": 12.753846153846155,
      "grad_norm": 1.53125,
      "learning_rate": 2.4673709896390938e-05,
      "loss": 0.0179,
      "step": 829
    },
    {
      "epoch": 12.76923076923077,
      "grad_norm": 3.953125,
      "learning_rate": 2.467172605732843e-05,
      "loss": 0.0159,
      "step": 830
    },
    {
      "epoch": 12.784615384615385,
      "grad_norm": 3.328125,
      "learning_rate": 2.4669736285914505e-05,
      "loss": 0.0546,
      "step": 831
    },
    {
      "epoch": 12.8,
      "grad_norm": 2.6875,
      "learning_rate": 2.4667740583118952e-05,
      "loss": 0.049,
      "step": 832
    },
    {
      "epoch": 12.815384615384616,
      "grad_norm": 8.4375,
      "learning_rate": 2.466573894991445e-05,
      "loss": 0.0704,
      "step": 833
    },
    {
      "epoch": 12.830769230769231,
      "grad_norm": 2.25,
      "learning_rate": 2.4663731387276583e-05,
      "loss": 0.0174,
      "step": 834
    },
    {
      "epoch": 12.846153846153847,
      "grad_norm": 2.5,
      "learning_rate": 2.46617178961838e-05,
      "loss": 0.0676,
      "step": 835
    },
    {
      "epoch": 12.861538461538462,
      "grad_norm": 1.765625,
      "learning_rate": 2.465969847761746e-05,
      "loss": 0.012,
      "step": 836
    },
    {
      "epoch": 12.876923076923077,
      "grad_norm": 0.7109375,
      "learning_rate": 2.4657673132561797e-05,
      "loss": 0.0333,
      "step": 837
    },
    {
      "epoch": 12.892307692307693,
      "grad_norm": 1.1640625,
      "learning_rate": 2.4655641862003943e-05,
      "loss": 0.0336,
      "step": 838
    },
    {
      "epoch": 12.907692307692308,
      "grad_norm": 1.5703125,
      "learning_rate": 2.465360466693391e-05,
      "loss": 0.0128,
      "step": 839
    },
    {
      "epoch": 12.923076923076923,
      "grad_norm": 7.59375,
      "learning_rate": 2.465156154834461e-05,
      "loss": 0.0684,
      "step": 840
    },
    {
      "epoch": 12.938461538461539,
      "grad_norm": 4.9375,
      "learning_rate": 2.464951250723183e-05,
      "loss": 0.0286,
      "step": 841
    },
    {
      "epoch": 12.953846153846154,
      "grad_norm": 7.34375,
      "learning_rate": 2.4647457544594246e-05,
      "loss": 0.0352,
      "step": 842
    },
    {
      "epoch": 12.96923076923077,
      "grad_norm": 11.4375,
      "learning_rate": 2.4645396661433424e-05,
      "loss": 0.1163,
      "step": 843
    },
    {
      "epoch": 12.984615384615385,
      "grad_norm": 3.390625,
      "learning_rate": 2.4643329858753818e-05,
      "loss": 0.0255,
      "step": 844
    },
    {
      "epoch": 13.0,
      "grad_norm": 2.9375,
      "learning_rate": 2.464125713756275e-05,
      "loss": 0.0562,
      "step": 845
    },
    {
      "epoch": 13.015384615384615,
      "grad_norm": 9.0,
      "learning_rate": 2.4639178498870452e-05,
      "loss": 0.0592,
      "step": 846
    },
    {
      "epoch": 13.03076923076923,
      "grad_norm": 7.3125,
      "learning_rate": 2.463709394369002e-05,
      "loss": 0.0826,
      "step": 847
    },
    {
      "epoch": 13.046153846153846,
      "grad_norm": 1.03125,
      "learning_rate": 2.4635003473037444e-05,
      "loss": 0.0323,
      "step": 848
    },
    {
      "epoch": 13.061538461538461,
      "grad_norm": 1.484375,
      "learning_rate": 2.4632907087931593e-05,
      "loss": 0.055,
      "step": 849
    },
    {
      "epoch": 13.076923076923077,
      "grad_norm": 2.203125,
      "learning_rate": 2.4630804789394218e-05,
      "loss": 0.026,
      "step": 850
    },
    {
      "epoch": 13.092307692307692,
      "grad_norm": 0.55859375,
      "learning_rate": 2.4628696578449956e-05,
      "loss": 0.0114,
      "step": 851
    },
    {
      "epoch": 13.107692307692307,
      "grad_norm": 3.1875,
      "learning_rate": 2.4626582456126324e-05,
      "loss": 0.0248,
      "step": 852
    },
    {
      "epoch": 13.123076923076923,
      "grad_norm": 3.984375,
      "learning_rate": 2.4624462423453716e-05,
      "loss": 0.0399,
      "step": 853
    },
    {
      "epoch": 13.138461538461538,
      "grad_norm": 5.6875,
      "learning_rate": 2.4622336481465414e-05,
      "loss": 0.0506,
      "step": 854
    },
    {
      "epoch": 13.153846153846153,
      "grad_norm": 4.9375,
      "learning_rate": 2.4620204631197574e-05,
      "loss": 0.0629,
      "step": 855
    },
    {
      "epoch": 13.169230769230769,
      "grad_norm": 2.28125,
      "learning_rate": 2.4618066873689238e-05,
      "loss": 0.0167,
      "step": 856
    },
    {
      "epoch": 13.184615384615384,
      "grad_norm": 3.203125,
      "learning_rate": 2.4615923209982316e-05,
      "loss": 0.0562,
      "step": 857
    },
    {
      "epoch": 13.2,
      "grad_norm": 8.4375,
      "learning_rate": 2.4613773641121607e-05,
      "loss": 0.1323,
      "step": 858
    },
    {
      "epoch": 13.215384615384615,
      "grad_norm": 2.84375,
      "learning_rate": 2.461161816815479e-05,
      "loss": 0.0152,
      "step": 859
    },
    {
      "epoch": 13.23076923076923,
      "grad_norm": 7.3125,
      "learning_rate": 2.4609456792132407e-05,
      "loss": 0.0526,
      "step": 860
    },
    {
      "epoch": 13.246153846153845,
      "grad_norm": 4.5,
      "learning_rate": 2.4607289514107888e-05,
      "loss": 0.0177,
      "step": 861
    },
    {
      "epoch": 13.261538461538462,
      "grad_norm": 4.03125,
      "learning_rate": 2.4605116335137547e-05,
      "loss": 0.0641,
      "step": 862
    },
    {
      "epoch": 13.276923076923078,
      "grad_norm": 5.75,
      "learning_rate": 2.4602937256280558e-05,
      "loss": 0.0808,
      "step": 863
    },
    {
      "epoch": 13.292307692307693,
      "grad_norm": 0.73828125,
      "learning_rate": 2.4600752278598978e-05,
      "loss": 0.0239,
      "step": 864
    },
    {
      "epoch": 13.307692307692308,
      "grad_norm": 3.609375,
      "learning_rate": 2.4598561403157734e-05,
      "loss": 0.0442,
      "step": 865
    },
    {
      "epoch": 13.323076923076924,
      "grad_norm": 5.84375,
      "learning_rate": 2.4596364631024643e-05,
      "loss": 0.0614,
      "step": 866
    },
    {
      "epoch": 13.338461538461539,
      "grad_norm": 2.453125,
      "learning_rate": 2.4594161963270373e-05,
      "loss": 0.0381,
      "step": 867
    },
    {
      "epoch": 13.353846153846154,
      "grad_norm": 3.375,
      "learning_rate": 2.459195340096849e-05,
      "loss": 0.0363,
      "step": 868
    },
    {
      "epoch": 13.36923076923077,
      "grad_norm": 1.984375,
      "learning_rate": 2.4589738945195413e-05,
      "loss": 0.0322,
      "step": 869
    },
    {
      "epoch": 13.384615384615385,
      "grad_norm": 2.578125,
      "learning_rate": 2.458751859703044e-05,
      "loss": 0.0317,
      "step": 870
    },
    {
      "epoch": 13.4,
      "grad_norm": 5.0,
      "learning_rate": 2.4585292357555746e-05,
      "loss": 0.0371,
      "step": 871
    },
    {
      "epoch": 13.415384615384616,
      "grad_norm": 5.5,
      "learning_rate": 2.4583060227856368e-05,
      "loss": 0.0803,
      "step": 872
    },
    {
      "epoch": 13.430769230769231,
      "grad_norm": 2.109375,
      "learning_rate": 2.458082220902022e-05,
      "loss": 0.0183,
      "step": 873
    },
    {
      "epoch": 13.446153846153846,
      "grad_norm": 5.34375,
      "learning_rate": 2.4578578302138087e-05,
      "loss": 0.0408,
      "step": 874
    },
    {
      "epoch": 13.461538461538462,
      "grad_norm": 4.90625,
      "learning_rate": 2.457632850830363e-05,
      "loss": 0.0608,
      "step": 875
    },
    {
      "epoch": 13.476923076923077,
      "grad_norm": 2.84375,
      "learning_rate": 2.4574072828613354e-05,
      "loss": 0.0309,
      "step": 876
    },
    {
      "epoch": 13.492307692307692,
      "grad_norm": 0.5859375,
      "learning_rate": 2.457181126416666e-05,
      "loss": 0.0557,
      "step": 877
    },
    {
      "epoch": 13.507692307692308,
      "grad_norm": 5.75,
      "learning_rate": 2.456954381606581e-05,
      "loss": 0.0435,
      "step": 878
    },
    {
      "epoch": 13.523076923076923,
      "grad_norm": 5.40625,
      "learning_rate": 2.4567270485415926e-05,
      "loss": 0.0422,
      "step": 879
    },
    {
      "epoch": 13.538461538461538,
      "grad_norm": 4.90625,
      "learning_rate": 2.4564991273325e-05,
      "loss": 0.0456,
      "step": 880
    },
    {
      "epoch": 13.553846153846154,
      "grad_norm": 2.921875,
      "learning_rate": 2.4562706180903894e-05,
      "loss": 0.0496,
      "step": 881
    },
    {
      "epoch": 13.569230769230769,
      "grad_norm": 3.4375,
      "learning_rate": 2.4560415209266334e-05,
      "loss": 0.043,
      "step": 882
    },
    {
      "epoch": 13.584615384615384,
      "grad_norm": 5.46875,
      "learning_rate": 2.455811835952892e-05,
      "loss": 0.0623,
      "step": 883
    },
    {
      "epoch": 13.6,
      "grad_norm": 7.71875,
      "learning_rate": 2.4555815632811092e-05,
      "loss": 0.1131,
      "step": 884
    },
    {
      "epoch": 13.615384615384615,
      "grad_norm": 6.25,
      "learning_rate": 2.4553507030235184e-05,
      "loss": 0.0615,
      "step": 885
    },
    {
      "epoch": 13.63076923076923,
      "grad_norm": 2.90625,
      "learning_rate": 2.455119255292638e-05,
      "loss": 0.0384,
      "step": 886
    },
    {
      "epoch": 13.646153846153846,
      "grad_norm": 6.03125,
      "learning_rate": 2.4548872202012718e-05,
      "loss": 0.0524,
      "step": 887
    },
    {
      "epoch": 13.661538461538461,
      "grad_norm": 2.328125,
      "learning_rate": 2.4546545978625126e-05,
      "loss": 0.0313,
      "step": 888
    },
    {
      "epoch": 13.676923076923076,
      "grad_norm": 4.15625,
      "learning_rate": 2.454421388389736e-05,
      "loss": 0.0496,
      "step": 889
    },
    {
      "epoch": 13.692307692307692,
      "grad_norm": 4.125,
      "learning_rate": 2.4541875918966063e-05,
      "loss": 0.0432,
      "step": 890
    },
    {
      "epoch": 13.707692307692307,
      "grad_norm": 1.890625,
      "learning_rate": 2.453953208497073e-05,
      "loss": 0.0279,
      "step": 891
    },
    {
      "epoch": 13.723076923076922,
      "grad_norm": 2.09375,
      "learning_rate": 2.4537182383053713e-05,
      "loss": 0.1034,
      "step": 892
    },
    {
      "epoch": 13.73846153846154,
      "grad_norm": 2.40625,
      "learning_rate": 2.4534826814360236e-05,
      "loss": 0.0523,
      "step": 893
    },
    {
      "epoch": 13.753846153846155,
      "grad_norm": 2.421875,
      "learning_rate": 2.453246538003837e-05,
      "loss": 0.0445,
      "step": 894
    },
    {
      "epoch": 13.76923076923077,
      "grad_norm": 2.46875,
      "learning_rate": 2.453009808123905e-05,
      "loss": 0.0254,
      "step": 895
    },
    {
      "epoch": 13.784615384615385,
      "grad_norm": 3.109375,
      "learning_rate": 2.452772491911607e-05,
      "loss": 0.0652,
      "step": 896
    },
    {
      "epoch": 13.8,
      "grad_norm": 2.53125,
      "learning_rate": 2.4525345894826073e-05,
      "loss": 0.0467,
      "step": 897
    },
    {
      "epoch": 13.815384615384616,
      "grad_norm": 2.96875,
      "learning_rate": 2.4522961009528577e-05,
      "loss": 0.0522,
      "step": 898
    },
    {
      "epoch": 13.830769230769231,
      "grad_norm": 4.125,
      "learning_rate": 2.452057026438594e-05,
      "loss": 0.0732,
      "step": 899
    },
    {
      "epoch": 13.846153846153847,
      "grad_norm": 3.703125,
      "learning_rate": 2.451817366056339e-05,
      "loss": 0.0481,
      "step": 900
    },
    {
      "epoch": 13.861538461538462,
      "grad_norm": 7.84375,
      "learning_rate": 2.4515771199228987e-05,
      "loss": 0.0599,
      "step": 901
    },
    {
      "epoch": 13.876923076923077,
      "grad_norm": 1.0625,
      "learning_rate": 2.4513362881553674e-05,
      "loss": 0.0185,
      "step": 902
    },
    {
      "epoch": 13.892307692307693,
      "grad_norm": 2.84375,
      "learning_rate": 2.4510948708711234e-05,
      "loss": 0.0229,
      "step": 903
    },
    {
      "epoch": 13.907692307692308,
      "grad_norm": 0.49609375,
      "learning_rate": 2.45085286818783e-05,
      "loss": 0.0451,
      "step": 904
    },
    {
      "epoch": 13.923076923076923,
      "grad_norm": 1.0,
      "learning_rate": 2.4506102802234372e-05,
      "loss": 0.0165,
      "step": 905
    },
    {
      "epoch": 13.938461538461539,
      "grad_norm": 4.9375,
      "learning_rate": 2.450367107096179e-05,
      "loss": 0.0384,
      "step": 906
    },
    {
      "epoch": 13.953846153846154,
      "grad_norm": 8.6875,
      "learning_rate": 2.450123348924575e-05,
      "loss": 0.1074,
      "step": 907
    },
    {
      "epoch": 13.96923076923077,
      "grad_norm": 3.546875,
      "learning_rate": 2.4498790058274297e-05,
      "loss": 0.0245,
      "step": 908
    },
    {
      "epoch": 13.984615384615385,
      "grad_norm": 4.78125,
      "learning_rate": 2.4496340779238335e-05,
      "loss": 0.0512,
      "step": 909
    },
    {
      "epoch": 14.0,
      "grad_norm": 3.625,
      "learning_rate": 2.449388565333161e-05,
      "loss": 0.0264,
      "step": 910
    },
    {
      "epoch": 14.015384615384615,
      "grad_norm": 2.765625,
      "learning_rate": 2.449142468175072e-05,
      "loss": 0.0342,
      "step": 911
    },
    {
      "epoch": 14.03076923076923,
      "grad_norm": 4.5,
      "learning_rate": 2.4488957865695114e-05,
      "loss": 0.0556,
      "step": 912
    },
    {
      "epoch": 14.046153846153846,
      "grad_norm": 1.0546875,
      "learning_rate": 2.4486485206367088e-05,
      "loss": 0.0645,
      "step": 913
    },
    {
      "epoch": 14.061538461538461,
      "grad_norm": 1.9375,
      "learning_rate": 2.4484006704971784e-05,
      "loss": 0.0231,
      "step": 914
    },
    {
      "epoch": 14.076923076923077,
      "grad_norm": 4.03125,
      "learning_rate": 2.44815223627172e-05,
      "loss": 0.0479,
      "step": 915
    },
    {
      "epoch": 14.092307692307692,
      "grad_norm": 6.84375,
      "learning_rate": 2.4479032180814166e-05,
      "loss": 0.0471,
      "step": 916
    },
    {
      "epoch": 14.107692307692307,
      "grad_norm": 3.484375,
      "learning_rate": 2.4476536160476372e-05,
      "loss": 0.0471,
      "step": 917
    },
    {
      "epoch": 14.123076923076923,
      "grad_norm": 5.1875,
      "learning_rate": 2.4474034302920347e-05,
      "loss": 0.0813,
      "step": 918
    },
    {
      "epoch": 14.138461538461538,
      "grad_norm": 5.625,
      "learning_rate": 2.447152660936546e-05,
      "loss": 0.0509,
      "step": 919
    },
    {
      "epoch": 14.153846153846153,
      "grad_norm": 3.953125,
      "learning_rate": 2.4469013081033936e-05,
      "loss": 0.0429,
      "step": 920
    },
    {
      "epoch": 14.169230769230769,
      "grad_norm": 10.375,
      "learning_rate": 2.446649371915084e-05,
      "loss": 0.0489,
      "step": 921
    },
    {
      "epoch": 14.184615384615384,
      "grad_norm": 1.6875,
      "learning_rate": 2.446396852494408e-05,
      "loss": 0.0481,
      "step": 922
    },
    {
      "epoch": 14.2,
      "grad_norm": 4.9375,
      "learning_rate": 2.4461437499644394e-05,
      "loss": 0.0425,
      "step": 923
    },
    {
      "epoch": 14.215384615384615,
      "grad_norm": 2.625,
      "learning_rate": 2.4458900644485385e-05,
      "loss": 0.0199,
      "step": 924
    },
    {
      "epoch": 14.23076923076923,
      "grad_norm": 3.453125,
      "learning_rate": 2.4456357960703475e-05,
      "loss": 0.0346,
      "step": 925
    },
    {
      "epoch": 14.246153846153845,
      "grad_norm": 5.0,
      "learning_rate": 2.4453809449537947e-05,
      "loss": 0.0295,
      "step": 926
    },
    {
      "epoch": 14.261538461538462,
      "grad_norm": 5.625,
      "learning_rate": 2.4451255112230904e-05,
      "loss": 0.0353,
      "step": 927
    },
    {
      "epoch": 14.276923076923078,
      "grad_norm": 2.390625,
      "learning_rate": 2.444869495002731e-05,
      "loss": 0.0407,
      "step": 928
    },
    {
      "epoch": 14.292307692307693,
      "grad_norm": 0.9375,
      "learning_rate": 2.4446128964174954e-05,
      "loss": 0.0221,
      "step": 929
    },
    {
      "epoch": 14.307692307692308,
      "grad_norm": 3.5,
      "learning_rate": 2.4443557155924457e-05,
      "loss": 0.022,
      "step": 930
    },
    {
      "epoch": 14.323076923076924,
      "grad_norm": 6.125,
      "learning_rate": 2.4440979526529295e-05,
      "loss": 0.0474,
      "step": 931
    },
    {
      "epoch": 14.338461538461539,
      "grad_norm": 0.9609375,
      "learning_rate": 2.4438396077245778e-05,
      "loss": 0.065,
      "step": 932
    },
    {
      "epoch": 14.353846153846154,
      "grad_norm": 4.40625,
      "learning_rate": 2.443580680933304e-05,
      "loss": 0.0696,
      "step": 933
    },
    {
      "epoch": 14.36923076923077,
      "grad_norm": 7.15625,
      "learning_rate": 2.4433211724053055e-05,
      "loss": 0.0442,
      "step": 934
    },
    {
      "epoch": 14.384615384615385,
      "grad_norm": 4.75,
      "learning_rate": 2.4430610822670654e-05,
      "loss": 0.0684,
      "step": 935
    },
    {
      "epoch": 14.4,
      "grad_norm": 5.25,
      "learning_rate": 2.4428004106453462e-05,
      "loss": 0.0292,
      "step": 936
    },
    {
      "epoch": 14.415384615384616,
      "grad_norm": 5.71875,
      "learning_rate": 2.4425391576671974e-05,
      "loss": 0.0294,
      "step": 937
    },
    {
      "epoch": 14.430769230769231,
      "grad_norm": 3.9375,
      "learning_rate": 2.4422773234599506e-05,
      "loss": 0.0596,
      "step": 938
    },
    {
      "epoch": 14.446153846153846,
      "grad_norm": 3.46875,
      "learning_rate": 2.44201490815122e-05,
      "loss": 0.0258,
      "step": 939
    },
    {
      "epoch": 14.461538461538462,
      "grad_norm": 7.5,
      "learning_rate": 2.441751911868904e-05,
      "loss": 0.0906,
      "step": 940
    },
    {
      "epoch": 14.476923076923077,
      "grad_norm": 0.85546875,
      "learning_rate": 2.4414883347411836e-05,
      "loss": 0.0134,
      "step": 941
    },
    {
      "epoch": 14.492307692307692,
      "grad_norm": 6.1875,
      "learning_rate": 2.441224176896524e-05,
      "loss": 0.0484,
      "step": 942
    },
    {
      "epoch": 14.507692307692308,
      "grad_norm": 4.5,
      "learning_rate": 2.4409594384636707e-05,
      "loss": 0.0406,
      "step": 943
    },
    {
      "epoch": 14.523076923076923,
      "grad_norm": 3.484375,
      "learning_rate": 2.4406941195716555e-05,
      "loss": 0.0533,
      "step": 944
    },
    {
      "epoch": 14.538461538461538,
      "grad_norm": 3.15625,
      "learning_rate": 2.440428220349791e-05,
      "loss": 0.0307,
      "step": 945
    },
    {
      "epoch": 14.553846153846154,
      "grad_norm": 2.109375,
      "learning_rate": 2.4401617409276735e-05,
      "loss": 0.0138,
      "step": 946
    },
    {
      "epoch": 14.569230769230769,
      "grad_norm": 3.640625,
      "learning_rate": 2.4398946814351818e-05,
      "loss": 0.0453,
      "step": 947
    },
    {
      "epoch": 14.584615384615384,
      "grad_norm": 2.609375,
      "learning_rate": 2.4396270420024774e-05,
      "loss": 0.0401,
      "step": 948
    },
    {
      "epoch": 14.6,
      "grad_norm": 10.0,
      "learning_rate": 2.439358822760004e-05,
      "loss": 0.0426,
      "step": 949
    },
    {
      "epoch": 14.615384615384615,
      "grad_norm": 3.921875,
      "learning_rate": 2.439090023838489e-05,
      "loss": 0.0679,
      "step": 950
    },
    {
      "epoch": 14.63076923076923,
      "grad_norm": 6.96875,
      "learning_rate": 2.438820645368942e-05,
      "loss": 0.1103,
      "step": 951
    },
    {
      "epoch": 14.646153846153846,
      "grad_norm": 1.1484375,
      "learning_rate": 2.438550687482654e-05,
      "loss": 0.0235,
      "step": 952
    },
    {
      "epoch": 14.661538461538461,
      "grad_norm": 2.4375,
      "learning_rate": 2.4382801503112003e-05,
      "loss": 0.0259,
      "step": 953
    },
    {
      "epoch": 14.676923076923076,
      "grad_norm": 2.359375,
      "learning_rate": 2.438009033986436e-05,
      "loss": 0.0228,
      "step": 954
    },
    {
      "epoch": 14.692307692307692,
      "grad_norm": 1.859375,
      "learning_rate": 2.437737338640501e-05,
      "loss": 0.015,
      "step": 955
    },
    {
      "epoch": 14.707692307692307,
      "grad_norm": 2.75,
      "learning_rate": 2.4374650644058156e-05,
      "loss": 0.035,
      "step": 956
    },
    {
      "epoch": 14.723076923076922,
      "grad_norm": 3.015625,
      "learning_rate": 2.4371922114150835e-05,
      "loss": 0.0513,
      "step": 957
    },
    {
      "epoch": 14.73846153846154,
      "grad_norm": 2.46875,
      "learning_rate": 2.43691877980129e-05,
      "loss": 0.0329,
      "step": 958
    },
    {
      "epoch": 14.753846153846155,
      "grad_norm": 3.953125,
      "learning_rate": 2.4366447696977018e-05,
      "loss": 0.0368,
      "step": 959
    },
    {
      "epoch": 14.76923076923077,
      "grad_norm": 6.28125,
      "learning_rate": 2.4363701812378683e-05,
      "loss": 0.0854,
      "step": 960
    },
    {
      "epoch": 14.784615384615385,
      "grad_norm": 2.40625,
      "learning_rate": 2.4360950145556208e-05,
      "loss": 0.0141,
      "step": 961
    },
    {
      "epoch": 14.8,
      "grad_norm": 2.859375,
      "learning_rate": 2.4358192697850718e-05,
      "loss": 0.0565,
      "step": 962
    },
    {
      "epoch": 14.815384615384616,
      "grad_norm": 4.21875,
      "learning_rate": 2.4355429470606166e-05,
      "loss": 0.0694,
      "step": 963
    },
    {
      "epoch": 14.830769230769231,
      "grad_norm": 1.4609375,
      "learning_rate": 2.4352660465169307e-05,
      "loss": 0.0273,
      "step": 964
    },
    {
      "epoch": 14.846153846153847,
      "grad_norm": 2.75,
      "learning_rate": 2.4349885682889727e-05,
      "loss": 0.0317,
      "step": 965
    },
    {
      "epoch": 14.861538461538462,
      "grad_norm": 3.296875,
      "learning_rate": 2.4347105125119824e-05,
      "loss": 0.0129,
      "step": 966
    },
    {
      "epoch": 14.876923076923077,
      "grad_norm": 3.703125,
      "learning_rate": 2.4344318793214797e-05,
      "loss": 0.0199,
      "step": 967
    },
    {
      "epoch": 14.892307692307693,
      "grad_norm": 5.6875,
      "learning_rate": 2.4341526688532678e-05,
      "loss": 0.0784,
      "step": 968
    },
    {
      "epoch": 14.907692307692308,
      "grad_norm": 6.40625,
      "learning_rate": 2.4338728812434305e-05,
      "loss": 0.102,
      "step": 969
    },
    {
      "epoch": 14.923076923076923,
      "grad_norm": 3.15625,
      "learning_rate": 2.4335925166283325e-05,
      "loss": 0.0234,
      "step": 970
    },
    {
      "epoch": 14.938461538461539,
      "grad_norm": 1.0078125,
      "learning_rate": 2.4333115751446208e-05,
      "loss": 0.0198,
      "step": 971
    },
    {
      "epoch": 14.953846153846154,
      "grad_norm": 2.921875,
      "learning_rate": 2.433030056929222e-05,
      "loss": 0.0739,
      "step": 972
    },
    {
      "epoch": 14.96923076923077,
      "grad_norm": 3.359375,
      "learning_rate": 2.4327479621193448e-05,
      "loss": 0.0295,
      "step": 973
    },
    {
      "epoch": 14.984615384615385,
      "grad_norm": 2.046875,
      "learning_rate": 2.4324652908524798e-05,
      "loss": 0.0389,
      "step": 974
    },
    {
      "epoch": 15.0,
      "grad_norm": 3.203125,
      "learning_rate": 2.432182043266396e-05,
      "loss": 0.0319,
      "step": 975
    },
    {
      "epoch": 15.015384615384615,
      "grad_norm": 2.25,
      "learning_rate": 2.4318982194991463e-05,
      "loss": 0.0239,
      "step": 976
    },
    {
      "epoch": 15.03076923076923,
      "grad_norm": 3.515625,
      "learning_rate": 2.431613819689062e-05,
      "loss": 0.068,
      "step": 977
    },
    {
      "epoch": 15.046153846153846,
      "grad_norm": 0.828125,
      "learning_rate": 2.431328843974756e-05,
      "loss": 0.0248,
      "step": 978
    },
    {
      "epoch": 15.061538461538461,
      "grad_norm": 8.125,
      "learning_rate": 2.4310432924951228e-05,
      "loss": 0.0879,
      "step": 979
    },
    {
      "epoch": 15.076923076923077,
      "grad_norm": 3.21875,
      "learning_rate": 2.4307571653893364e-05,
      "loss": 0.064,
      "step": 980
    },
    {
      "epoch": 15.092307692307692,
      "grad_norm": 2.40625,
      "learning_rate": 2.4304704627968515e-05,
      "loss": 0.0439,
      "step": 981
    },
    {
      "epoch": 15.107692307692307,
      "grad_norm": 0.78125,
      "learning_rate": 2.4301831848574035e-05,
      "loss": 0.019,
      "step": 982
    },
    {
      "epoch": 15.123076923076923,
      "grad_norm": 3.421875,
      "learning_rate": 2.429895331711008e-05,
      "loss": 0.0445,
      "step": 983
    },
    {
      "epoch": 15.138461538461538,
      "grad_norm": 4.34375,
      "learning_rate": 2.429606903497962e-05,
      "loss": 0.0452,
      "step": 984
    },
    {
      "epoch": 15.153846153846153,
      "grad_norm": 3.0625,
      "learning_rate": 2.4293179003588407e-05,
      "loss": 0.0148,
      "step": 985
    },
    {
      "epoch": 15.169230769230769,
      "grad_norm": 2.734375,
      "learning_rate": 2.429028322434501e-05,
      "loss": 0.0224,
      "step": 986
    },
    {
      "epoch": 15.184615384615384,
      "grad_norm": 0.419921875,
      "learning_rate": 2.4287381698660807e-05,
      "loss": 0.0148,
      "step": 987
    },
    {
      "epoch": 15.2,
      "grad_norm": 3.5625,
      "learning_rate": 2.4284474427949956e-05,
      "loss": 0.0332,
      "step": 988
    },
    {
      "epoch": 15.215384615384615,
      "grad_norm": 8.4375,
      "learning_rate": 2.428156141362943e-05,
      "loss": 0.0343,
      "step": 989
    },
    {
      "epoch": 15.23076923076923,
      "grad_norm": 0.64453125,
      "learning_rate": 2.427864265711899e-05,
      "loss": 0.039,
      "step": 990
    },
    {
      "epoch": 15.246153846153845,
      "grad_norm": 2.015625,
      "learning_rate": 2.427571815984121e-05,
      "loss": 0.0348,
      "step": 991
    },
    {
      "epoch": 15.261538461538462,
      "grad_norm": 3.828125,
      "learning_rate": 2.427278792322145e-05,
      "loss": 0.0146,
      "step": 992
    },
    {
      "epoch": 15.276923076923078,
      "grad_norm": 2.625,
      "learning_rate": 2.4269851948687877e-05,
      "loss": 0.0737,
      "step": 993
    },
    {
      "epoch": 15.292307692307693,
      "grad_norm": 0.515625,
      "learning_rate": 2.426691023767145e-05,
      "loss": 0.0092,
      "step": 994
    },
    {
      "epoch": 15.307692307692308,
      "grad_norm": 2.90625,
      "learning_rate": 2.426396279160591e-05,
      "loss": 0.0158,
      "step": 995
    },
    {
      "epoch": 15.323076923076924,
      "grad_norm": 0.1806640625,
      "learning_rate": 2.426100961192782e-05,
      "loss": 0.0179,
      "step": 996
    },
    {
      "epoch": 15.338461538461539,
      "grad_norm": 15.75,
      "learning_rate": 2.425805070007652e-05,
      "loss": 0.0711,
      "step": 997
    },
    {
      "epoch": 15.353846153846154,
      "grad_norm": 3.53125,
      "learning_rate": 2.4255086057494146e-05,
      "loss": 0.0308,
      "step": 998
    },
    {
      "epoch": 15.36923076923077,
      "grad_norm": 4.9375,
      "learning_rate": 2.4252115685625627e-05,
      "loss": 0.0205,
      "step": 999
    },
    {
      "epoch": 15.384615384615385,
      "grad_norm": 6.53125,
      "learning_rate": 2.4249139585918695e-05,
      "loss": 0.0886,
      "step": 1000
    },
    {
      "epoch": 15.4,
      "grad_norm": 6.21875,
      "learning_rate": 2.4246157759823855e-05,
      "loss": 0.1021,
      "step": 1001
    },
    {
      "epoch": 15.415384615384616,
      "grad_norm": 5.34375,
      "learning_rate": 2.4243170208794418e-05,
      "loss": 0.0721,
      "step": 1002
    },
    {
      "epoch": 15.430769230769231,
      "grad_norm": 0.9453125,
      "learning_rate": 2.424017693428648e-05,
      "loss": 0.01,
      "step": 1003
    },
    {
      "epoch": 15.446153846153846,
      "grad_norm": 6.125,
      "learning_rate": 2.4237177937758927e-05,
      "loss": 0.0586,
      "step": 1004
    },
    {
      "epoch": 15.461538461538462,
      "grad_norm": 4.375,
      "learning_rate": 2.423417322067343e-05,
      "loss": 0.0507,
      "step": 1005
    },
    {
      "epoch": 15.476923076923077,
      "grad_norm": 7.6875,
      "learning_rate": 2.423116278449445e-05,
      "loss": 0.0534,
      "step": 1006
    },
    {
      "epoch": 15.492307692307692,
      "grad_norm": 2.859375,
      "learning_rate": 2.4228146630689245e-05,
      "loss": 0.0373,
      "step": 1007
    },
    {
      "epoch": 15.507692307692308,
      "grad_norm": 4.8125,
      "learning_rate": 2.4225124760727847e-05,
      "loss": 0.0593,
      "step": 1008
    },
    {
      "epoch": 15.523076923076923,
      "grad_norm": 2.8125,
      "learning_rate": 2.4222097176083074e-05,
      "loss": 0.0364,
      "step": 1009
    },
    {
      "epoch": 15.538461538461538,
      "grad_norm": 3.46875,
      "learning_rate": 2.421906387823055e-05,
      "loss": 0.0304,
      "step": 1010
    },
    {
      "epoch": 15.553846153846154,
      "grad_norm": 2.796875,
      "learning_rate": 2.4216024868648644e-05,
      "loss": 0.0372,
      "step": 1011
    },
    {
      "epoch": 15.569230769230769,
      "grad_norm": 2.296875,
      "learning_rate": 2.421298014881855e-05,
      "loss": 0.0513,
      "step": 1012
    },
    {
      "epoch": 15.584615384615384,
      "grad_norm": 1.0,
      "learning_rate": 2.4209929720224223e-05,
      "loss": 0.0431,
      "step": 1013
    },
    {
      "epoch": 15.6,
      "grad_norm": 2.578125,
      "learning_rate": 2.4206873584352404e-05,
      "loss": 0.0349,
      "step": 1014
    },
    {
      "epoch": 15.615384615384615,
      "grad_norm": 6.375,
      "learning_rate": 2.4203811742692617e-05,
      "loss": 0.0609,
      "step": 1015
    },
    {
      "epoch": 15.63076923076923,
      "grad_norm": 6.40625,
      "learning_rate": 2.420074419673717e-05,
      "loss": 0.0559,
      "step": 1016
    },
    {
      "epoch": 15.646153846153846,
      "grad_norm": 3.296875,
      "learning_rate": 2.419767094798114e-05,
      "loss": 0.0222,
      "step": 1017
    },
    {
      "epoch": 15.661538461538461,
      "grad_norm": 1.7421875,
      "learning_rate": 2.4194591997922393e-05,
      "loss": 0.048,
      "step": 1018
    },
    {
      "epoch": 15.676923076923076,
      "grad_norm": 1.8828125,
      "learning_rate": 2.4191507348061572e-05,
      "loss": 0.0257,
      "step": 1019
    },
    {
      "epoch": 15.692307692307692,
      "grad_norm": 2.328125,
      "learning_rate": 2.4188416999902104e-05,
      "loss": 0.0316,
      "step": 1020
    },
    {
      "epoch": 15.707692307692307,
      "grad_norm": 2.796875,
      "learning_rate": 2.418532095495018e-05,
      "loss": 0.0494,
      "step": 1021
    },
    {
      "epoch": 15.723076923076922,
      "grad_norm": 2.578125,
      "learning_rate": 2.418221921471478e-05,
      "loss": 0.0294,
      "step": 1022
    },
    {
      "epoch": 15.73846153846154,
      "grad_norm": 1.7734375,
      "learning_rate": 2.4179111780707652e-05,
      "loss": 0.0246,
      "step": 1023
    },
    {
      "epoch": 15.753846153846155,
      "grad_norm": 2.6875,
      "learning_rate": 2.417599865444332e-05,
      "loss": 0.0748,
      "step": 1024
    },
    {
      "epoch": 15.76923076923077,
      "grad_norm": 2.078125,
      "learning_rate": 2.4172879837439086e-05,
      "loss": 0.0393,
      "step": 1025
    },
    {
      "epoch": 15.784615384615385,
      "grad_norm": 1.953125,
      "learning_rate": 2.4169755331215023e-05,
      "loss": 0.0154,
      "step": 1026
    },
    {
      "epoch": 15.8,
      "grad_norm": 4.75,
      "learning_rate": 2.4166625137293976e-05,
      "loss": 0.0373,
      "step": 1027
    },
    {
      "epoch": 15.815384615384616,
      "grad_norm": 2.34375,
      "learning_rate": 2.4163489257201564e-05,
      "loss": 0.0171,
      "step": 1028
    },
    {
      "epoch": 15.830769230769231,
      "grad_norm": 2.40625,
      "learning_rate": 2.4160347692466184e-05,
      "loss": 0.0275,
      "step": 1029
    },
    {
      "epoch": 15.846153846153847,
      "grad_norm": 1.609375,
      "learning_rate": 2.4157200444618985e-05,
      "loss": 0.0212,
      "step": 1030
    },
    {
      "epoch": 15.861538461538462,
      "grad_norm": 4.9375,
      "learning_rate": 2.4154047515193904e-05,
      "loss": 0.0463,
      "step": 1031
    },
    {
      "epoch": 15.876923076923077,
      "grad_norm": 1.8984375,
      "learning_rate": 2.4150888905727638e-05,
      "loss": 0.022,
      "step": 1032
    },
    {
      "epoch": 15.892307692307693,
      "grad_norm": 5.21875,
      "learning_rate": 2.4147724617759656e-05,
      "loss": 0.0342,
      "step": 1033
    },
    {
      "epoch": 15.907692307692308,
      "grad_norm": 2.59375,
      "learning_rate": 2.4144554652832196e-05,
      "loss": 0.059,
      "step": 1034
    },
    {
      "epoch": 15.923076923076923,
      "grad_norm": 1.9609375,
      "learning_rate": 2.4141379012490255e-05,
      "loss": 0.0171,
      "step": 1035
    },
    {
      "epoch": 15.938461538461539,
      "grad_norm": 1.890625,
      "learning_rate": 2.4138197698281606e-05,
      "loss": 0.0292,
      "step": 1036
    },
    {
      "epoch": 15.953846153846154,
      "grad_norm": 2.0625,
      "learning_rate": 2.4135010711756777e-05,
      "loss": 0.0192,
      "step": 1037
    },
    {
      "epoch": 15.96923076923077,
      "grad_norm": 9.0,
      "learning_rate": 2.4131818054469078e-05,
      "loss": 0.1023,
      "step": 1038
    },
    {
      "epoch": 15.984615384615385,
      "grad_norm": 2.734375,
      "learning_rate": 2.4128619727974562e-05,
      "loss": 0.0272,
      "step": 1039
    },
    {
      "epoch": 16.0,
      "grad_norm": 3.484375,
      "learning_rate": 2.412541573383205e-05,
      "loss": 0.0377,
      "step": 1040
    },
    {
      "epoch": 16.015384615384615,
      "grad_norm": 0.5078125,
      "learning_rate": 2.4122206073603142e-05,
      "loss": 0.0137,
      "step": 1041
    },
    {
      "epoch": 16.03076923076923,
      "grad_norm": 2.09375,
      "learning_rate": 2.4118990748852183e-05,
      "loss": 0.0177,
      "step": 1042
    },
    {
      "epoch": 16.046153846153846,
      "grad_norm": 4.0625,
      "learning_rate": 2.4115769761146282e-05,
      "loss": 0.0398,
      "step": 1043
    },
    {
      "epoch": 16.06153846153846,
      "grad_norm": 5.53125,
      "learning_rate": 2.4112543112055308e-05,
      "loss": 0.0444,
      "step": 1044
    },
    {
      "epoch": 16.076923076923077,
      "grad_norm": 1.53125,
      "learning_rate": 2.410931080315189e-05,
      "loss": 0.017,
      "step": 1045
    },
    {
      "epoch": 16.092307692307692,
      "grad_norm": 1.265625,
      "learning_rate": 2.4106072836011422e-05,
      "loss": 0.0187,
      "step": 1046
    },
    {
      "epoch": 16.107692307692307,
      "grad_norm": 3.578125,
      "learning_rate": 2.410282921221204e-05,
      "loss": 0.0389,
      "step": 1047
    },
    {
      "epoch": 16.123076923076923,
      "grad_norm": 10.625,
      "learning_rate": 2.409957993333466e-05,
      "loss": 0.0545,
      "step": 1048
    },
    {
      "epoch": 16.138461538461538,
      "grad_norm": 8.0,
      "learning_rate": 2.4096325000962927e-05,
      "loss": 0.0441,
      "step": 1049
    },
    {
      "epoch": 16.153846153846153,
      "grad_norm": 3.6875,
      "learning_rate": 2.409306441668327e-05,
      "loss": 0.0395,
      "step": 1050
    },
    {
      "epoch": 16.16923076923077,
      "grad_norm": 4.875,
      "learning_rate": 2.4089798182084845e-05,
      "loss": 0.0344,
      "step": 1051
    },
    {
      "epoch": 16.184615384615384,
      "grad_norm": 6.8125,
      "learning_rate": 2.408652629875958e-05,
      "loss": 0.0476,
      "step": 1052
    },
    {
      "epoch": 16.2,
      "grad_norm": 1.8671875,
      "learning_rate": 2.408324876830215e-05,
      "loss": 0.0528,
      "step": 1053
    },
    {
      "epoch": 16.215384615384615,
      "grad_norm": 2.109375,
      "learning_rate": 2.407996559230999e-05,
      "loss": 0.027,
      "step": 1054
    },
    {
      "epoch": 16.23076923076923,
      "grad_norm": 1.140625,
      "learning_rate": 2.4076676772383277e-05,
      "loss": 0.0149,
      "step": 1055
    },
    {
      "epoch": 16.246153846153845,
      "grad_norm": 4.1875,
      "learning_rate": 2.407338231012494e-05,
      "loss": 0.0592,
      "step": 1056
    },
    {
      "epoch": 16.26153846153846,
      "grad_norm": 1.890625,
      "learning_rate": 2.4070082207140653e-05,
      "loss": 0.0352,
      "step": 1057
    },
    {
      "epoch": 16.276923076923076,
      "grad_norm": 4.09375,
      "learning_rate": 2.4066776465038858e-05,
      "loss": 0.064,
      "step": 1058
    },
    {
      "epoch": 16.29230769230769,
      "grad_norm": 2.234375,
      "learning_rate": 2.4063465085430725e-05,
      "loss": 0.0588,
      "step": 1059
    },
    {
      "epoch": 16.307692307692307,
      "grad_norm": 3.890625,
      "learning_rate": 2.4060148069930188e-05,
      "loss": 0.033,
      "step": 1060
    },
    {
      "epoch": 16.323076923076922,
      "grad_norm": 0.8984375,
      "learning_rate": 2.4056825420153917e-05,
      "loss": 0.0234,
      "step": 1061
    },
    {
      "epoch": 16.338461538461537,
      "grad_norm": 5.40625,
      "learning_rate": 2.4053497137721324e-05,
      "loss": 0.0488,
      "step": 1062
    },
    {
      "epoch": 16.353846153846153,
      "grad_norm": 3.6875,
      "learning_rate": 2.4050163224254586e-05,
      "loss": 0.0212,
      "step": 1063
    },
    {
      "epoch": 16.369230769230768,
      "grad_norm": 3.53125,
      "learning_rate": 2.4046823681378603e-05,
      "loss": 0.0498,
      "step": 1064
    },
    {
      "epoch": 16.384615384615383,
      "grad_norm": 0.75390625,
      "learning_rate": 2.404347851072103e-05,
      "loss": 0.005,
      "step": 1065
    },
    {
      "epoch": 16.4,
      "grad_norm": 4.875,
      "learning_rate": 2.4040127713912264e-05,
      "loss": 0.0381,
      "step": 1066
    },
    {
      "epoch": 16.415384615384614,
      "grad_norm": 3.46875,
      "learning_rate": 2.4036771292585438e-05,
      "loss": 0.0362,
      "step": 1067
    },
    {
      "epoch": 16.43076923076923,
      "grad_norm": 3.390625,
      "learning_rate": 2.4033409248376437e-05,
      "loss": 0.0512,
      "step": 1068
    },
    {
      "epoch": 16.446153846153845,
      "grad_norm": 3.09375,
      "learning_rate": 2.4030041582923877e-05,
      "loss": 0.0811,
      "step": 1069
    },
    {
      "epoch": 16.46153846153846,
      "grad_norm": 2.03125,
      "learning_rate": 2.4026668297869118e-05,
      "loss": 0.0138,
      "step": 1070
    },
    {
      "epoch": 16.476923076923075,
      "grad_norm": 2.5,
      "learning_rate": 2.4023289394856257e-05,
      "loss": 0.0163,
      "step": 1071
    },
    {
      "epoch": 16.49230769230769,
      "grad_norm": 2.65625,
      "learning_rate": 2.4019904875532134e-05,
      "loss": 0.0658,
      "step": 1072
    },
    {
      "epoch": 16.50769230769231,
      "grad_norm": 2.546875,
      "learning_rate": 2.401651474154632e-05,
      "loss": 0.0199,
      "step": 1073
    },
    {
      "epoch": 16.523076923076925,
      "grad_norm": 1.5546875,
      "learning_rate": 2.401311899455112e-05,
      "loss": 0.0177,
      "step": 1074
    },
    {
      "epoch": 16.53846153846154,
      "grad_norm": 6.25,
      "learning_rate": 2.400971763620159e-05,
      "loss": 0.0365,
      "step": 1075
    },
    {
      "epoch": 16.553846153846155,
      "grad_norm": 7.5,
      "learning_rate": 2.4006310668155508e-05,
      "loss": 0.0474,
      "step": 1076
    },
    {
      "epoch": 16.56923076923077,
      "grad_norm": 3.5,
      "learning_rate": 2.400289809207338e-05,
      "loss": 0.0145,
      "step": 1077
    },
    {
      "epoch": 16.584615384615386,
      "grad_norm": 7.96875,
      "learning_rate": 2.3999479909618466e-05,
      "loss": 0.037,
      "step": 1078
    },
    {
      "epoch": 16.6,
      "grad_norm": 3.15625,
      "learning_rate": 2.399605612245674e-05,
      "loss": 0.0119,
      "step": 1079
    },
    {
      "epoch": 16.615384615384617,
      "grad_norm": 4.15625,
      "learning_rate": 2.3992626732256912e-05,
      "loss": 0.0543,
      "step": 1080
    },
    {
      "epoch": 16.630769230769232,
      "grad_norm": 4.25,
      "learning_rate": 2.398919174069043e-05,
      "loss": 0.0188,
      "step": 1081
    },
    {
      "epoch": 16.646153846153847,
      "grad_norm": 9.625,
      "learning_rate": 2.3985751149431468e-05,
      "loss": 0.0578,
      "step": 1082
    },
    {
      "epoch": 16.661538461538463,
      "grad_norm": 0.40234375,
      "learning_rate": 2.3982304960156925e-05,
      "loss": 0.0323,
      "step": 1083
    },
    {
      "epoch": 16.676923076923078,
      "grad_norm": 0.7578125,
      "learning_rate": 2.3978853174546433e-05,
      "loss": 0.0108,
      "step": 1084
    },
    {
      "epoch": 16.692307692307693,
      "grad_norm": 1.8515625,
      "learning_rate": 2.397539579428235e-05,
      "loss": 0.0088,
      "step": 1085
    },
    {
      "epoch": 16.70769230769231,
      "grad_norm": 13.25,
      "learning_rate": 2.3971932821049765e-05,
      "loss": 0.1008,
      "step": 1086
    },
    {
      "epoch": 16.723076923076924,
      "grad_norm": 2.1875,
      "learning_rate": 2.3968464256536482e-05,
      "loss": 0.0702,
      "step": 1087
    },
    {
      "epoch": 16.73846153846154,
      "grad_norm": 2.390625,
      "learning_rate": 2.3964990102433044e-05,
      "loss": 0.036,
      "step": 1088
    },
    {
      "epoch": 16.753846153846155,
      "grad_norm": 2.546875,
      "learning_rate": 2.3961510360432707e-05,
      "loss": 0.0232,
      "step": 1089
    },
    {
      "epoch": 16.76923076923077,
      "grad_norm": 5.84375,
      "learning_rate": 2.3958025032231455e-05,
      "loss": 0.0569,
      "step": 1090
    },
    {
      "epoch": 16.784615384615385,
      "grad_norm": 1.890625,
      "learning_rate": 2.3954534119527996e-05,
      "loss": 0.014,
      "step": 1091
    },
    {
      "epoch": 16.8,
      "grad_norm": 7.03125,
      "learning_rate": 2.3951037624023758e-05,
      "loss": 0.0395,
      "step": 1092
    },
    {
      "epoch": 16.815384615384616,
      "grad_norm": 2.265625,
      "learning_rate": 2.3947535547422896e-05,
      "loss": 0.0402,
      "step": 1093
    },
    {
      "epoch": 16.83076923076923,
      "grad_norm": 4.40625,
      "learning_rate": 2.394402789143227e-05,
      "loss": 0.0291,
      "step": 1094
    },
    {
      "epoch": 16.846153846153847,
      "grad_norm": 2.109375,
      "learning_rate": 2.394051465776147e-05,
      "loss": 0.0238,
      "step": 1095
    },
    {
      "epoch": 16.861538461538462,
      "grad_norm": 3.078125,
      "learning_rate": 2.3936995848122812e-05,
      "loss": 0.0387,
      "step": 1096
    },
    {
      "epoch": 16.876923076923077,
      "grad_norm": 2.75,
      "learning_rate": 2.3933471464231312e-05,
      "loss": 0.051,
      "step": 1097
    },
    {
      "epoch": 16.892307692307693,
      "grad_norm": 1.90625,
      "learning_rate": 2.392994150780472e-05,
      "loss": 0.0153,
      "step": 1098
    },
    {
      "epoch": 16.907692307692308,
      "grad_norm": 1.8828125,
      "learning_rate": 2.392640598056349e-05,
      "loss": 0.0279,
      "step": 1099
    },
    {
      "epoch": 16.923076923076923,
      "grad_norm": 1.328125,
      "learning_rate": 2.3922864884230788e-05,
      "loss": 0.0155,
      "step": 1100
    },
    {
      "epoch": 16.93846153846154,
      "grad_norm": 0.9453125,
      "learning_rate": 2.391931822053251e-05,
      "loss": 0.0246,
      "step": 1101
    },
    {
      "epoch": 16.953846153846154,
      "grad_norm": 2.28125,
      "learning_rate": 2.391576599119726e-05,
      "loss": 0.0285,
      "step": 1102
    },
    {
      "epoch": 16.96923076923077,
      "grad_norm": 2.6875,
      "learning_rate": 2.3912208197956343e-05,
      "loss": 0.0259,
      "step": 1103
    },
    {
      "epoch": 16.984615384615385,
      "grad_norm": 1.2265625,
      "learning_rate": 2.3908644842543788e-05,
      "loss": 0.0241,
      "step": 1104
    },
    {
      "epoch": 17.0,
      "grad_norm": 2.703125,
      "learning_rate": 2.390507592669633e-05,
      "loss": 0.0776,
      "step": 1105
    },
    {
      "epoch": 17.015384615384615,
      "grad_norm": 1.0625,
      "learning_rate": 2.390150145215341e-05,
      "loss": 0.0144,
      "step": 1106
    },
    {
      "epoch": 17.03076923076923,
      "grad_norm": 1.078125,
      "learning_rate": 2.38979214206572e-05,
      "loss": 0.011,
      "step": 1107
    },
    {
      "epoch": 17.046153846153846,
      "grad_norm": 4.0,
      "learning_rate": 2.3894335833952548e-05,
      "loss": 0.0311,
      "step": 1108
    },
    {
      "epoch": 17.06153846153846,
      "grad_norm": 1.5078125,
      "learning_rate": 2.389074469378703e-05,
      "loss": 0.0409,
      "step": 1109
    },
    {
      "epoch": 17.076923076923077,
      "grad_norm": 7.3125,
      "learning_rate": 2.3887148001910925e-05,
      "loss": 0.0768,
      "step": 1110
    },
    {
      "epoch": 17.092307692307692,
      "grad_norm": 5.625,
      "learning_rate": 2.3883545760077215e-05,
      "loss": 0.0849,
      "step": 1111
    },
    {
      "epoch": 17.107692307692307,
      "grad_norm": 0.94140625,
      "learning_rate": 2.3879937970041595e-05,
      "loss": 0.0073,
      "step": 1112
    },
    {
      "epoch": 17.123076923076923,
      "grad_norm": 3.828125,
      "learning_rate": 2.3876324633562446e-05,
      "loss": 0.0716,
      "step": 1113
    },
    {
      "epoch": 17.138461538461538,
      "grad_norm": 0.58203125,
      "learning_rate": 2.3872705752400874e-05,
      "loss": 0.0106,
      "step": 1114
    },
    {
      "epoch": 17.153846153846153,
      "grad_norm": 7.6875,
      "learning_rate": 2.386908132832068e-05,
      "loss": 0.0459,
      "step": 1115
    },
    {
      "epoch": 17.16923076923077,
      "grad_norm": 3.6875,
      "learning_rate": 2.386545136308836e-05,
      "loss": 0.0486,
      "step": 1116
    },
    {
      "epoch": 17.184615384615384,
      "grad_norm": 3.4375,
      "learning_rate": 2.386181585847311e-05,
      "loss": 0.0245,
      "step": 1117
    },
    {
      "epoch": 17.2,
      "grad_norm": 2.46875,
      "learning_rate": 2.3858174816246837e-05,
      "loss": 0.0672,
      "step": 1118
    },
    {
      "epoch": 17.215384615384615,
      "grad_norm": 6.9375,
      "learning_rate": 2.385452823818414e-05,
      "loss": 0.0185,
      "step": 1119
    },
    {
      "epoch": 17.23076923076923,
      "grad_norm": 4.375,
      "learning_rate": 2.3850876126062314e-05,
      "loss": 0.0321,
      "step": 1120
    },
    {
      "epoch": 17.246153846153845,
      "grad_norm": 3.109375,
      "learning_rate": 2.384721848166136e-05,
      "loss": 0.0416,
      "step": 1121
    },
    {
      "epoch": 17.26153846153846,
      "grad_norm": 6.4375,
      "learning_rate": 2.3843555306763963e-05,
      "loss": 0.061,
      "step": 1122
    },
    {
      "epoch": 17.276923076923076,
      "grad_norm": 1.1640625,
      "learning_rate": 2.3839886603155516e-05,
      "loss": 0.0212,
      "step": 1123
    },
    {
      "epoch": 17.29230769230769,
      "grad_norm": 5.15625,
      "learning_rate": 2.3836212372624088e-05,
      "loss": 0.0609,
      "step": 1124
    },
    {
      "epoch": 17.307692307692307,
      "grad_norm": 2.921875,
      "learning_rate": 2.3832532616960473e-05,
      "loss": 0.0132,
      "step": 1125
    },
    {
      "epoch": 17.323076923076922,
      "grad_norm": 0.1689453125,
      "learning_rate": 2.3828847337958126e-05,
      "loss": 0.009,
      "step": 1126
    },
    {
      "epoch": 17.338461538461537,
      "grad_norm": 2.171875,
      "learning_rate": 2.3825156537413213e-05,
      "loss": 0.0254,
      "step": 1127
    },
    {
      "epoch": 17.353846153846153,
      "grad_norm": 4.125,
      "learning_rate": 2.382146021712458e-05,
      "loss": 0.0263,
      "step": 1128
    },
    {
      "epoch": 17.369230769230768,
      "grad_norm": 7.28125,
      "learning_rate": 2.3817758378893774e-05,
      "loss": 0.0269,
      "step": 1129
    },
    {
      "epoch": 17.384615384615383,
      "grad_norm": 11.5625,
      "learning_rate": 2.3814051024525018e-05,
      "loss": 0.0441,
      "step": 1130
    },
    {
      "epoch": 17.4,
      "grad_norm": 0.91796875,
      "learning_rate": 2.3810338155825245e-05,
      "loss": 0.0187,
      "step": 1131
    },
    {
      "epoch": 17.415384615384614,
      "grad_norm": 1.578125,
      "learning_rate": 2.3806619774604053e-05,
      "loss": 0.0085,
      "step": 1132
    },
    {
      "epoch": 17.43076923076923,
      "grad_norm": 3.03125,
      "learning_rate": 2.3802895882673734e-05,
      "loss": 0.0259,
      "step": 1133
    },
    {
      "epoch": 17.446153846153845,
      "grad_norm": 0.5703125,
      "learning_rate": 2.3799166481849272e-05,
      "loss": 0.0085,
      "step": 1134
    },
    {
      "epoch": 17.46153846153846,
      "grad_norm": 6.78125,
      "learning_rate": 2.3795431573948333e-05,
      "loss": 0.0497,
      "step": 1135
    },
    {
      "epoch": 17.476923076923075,
      "grad_norm": 10.6875,
      "learning_rate": 2.379169116079126e-05,
      "loss": 0.0748,
      "step": 1136
    },
    {
      "epoch": 17.49230769230769,
      "grad_norm": 0.458984375,
      "learning_rate": 2.3787945244201092e-05,
      "loss": 0.0063,
      "step": 1137
    },
    {
      "epoch": 17.50769230769231,
      "grad_norm": 4.40625,
      "learning_rate": 2.3784193826003538e-05,
      "loss": 0.0962,
      "step": 1138
    },
    {
      "epoch": 17.523076923076925,
      "grad_norm": 0.97265625,
      "learning_rate": 2.3780436908026992e-05,
      "loss": 0.0404,
      "step": 1139
    },
    {
      "epoch": 17.53846153846154,
      "grad_norm": 1.1796875,
      "learning_rate": 2.3776674492102535e-05,
      "loss": 0.0681,
      "step": 1140
    },
    {
      "epoch": 17.553846153846155,
      "grad_norm": 0.53125,
      "learning_rate": 2.3772906580063924e-05,
      "loss": 0.0061,
      "step": 1141
    },
    {
      "epoch": 17.56923076923077,
      "grad_norm": 6.59375,
      "learning_rate": 2.376913317374759e-05,
      "loss": 0.0323,
      "step": 1142
    },
    {
      "epoch": 17.584615384615386,
      "grad_norm": 1.1015625,
      "learning_rate": 2.3765354274992643e-05,
      "loss": 0.0184,
      "step": 1143
    },
    {
      "epoch": 17.6,
      "grad_norm": 6.03125,
      "learning_rate": 2.3761569885640877e-05,
      "loss": 0.0691,
      "step": 1144
    },
    {
      "epoch": 17.615384615384617,
      "grad_norm": 3.4375,
      "learning_rate": 2.3757780007536757e-05,
      "loss": 0.017,
      "step": 1145
    },
    {
      "epoch": 17.630769230769232,
      "grad_norm": 4.3125,
      "learning_rate": 2.3753984642527423e-05,
      "loss": 0.0432,
      "step": 1146
    },
    {
      "epoch": 17.646153846153847,
      "grad_norm": 0.5078125,
      "learning_rate": 2.3750183792462686e-05,
      "loss": 0.0118,
      "step": 1147
    },
    {
      "epoch": 17.661538461538463,
      "grad_norm": 6.0,
      "learning_rate": 2.3746377459195037e-05,
      "loss": 0.035,
      "step": 1148
    },
    {
      "epoch": 17.676923076923078,
      "grad_norm": 2.3125,
      "learning_rate": 2.3742565644579638e-05,
      "loss": 0.0306,
      "step": 1149
    },
    {
      "epoch": 17.692307692307693,
      "grad_norm": 3.71875,
      "learning_rate": 2.373874835047432e-05,
      "loss": 0.023,
      "step": 1150
    },
    {
      "epoch": 17.70769230769231,
      "grad_norm": 1.7734375,
      "learning_rate": 2.3734925578739588e-05,
      "loss": 0.0195,
      "step": 1151
    },
    {
      "epoch": 17.723076923076924,
      "grad_norm": 3.984375,
      "learning_rate": 2.373109733123861e-05,
      "loss": 0.0481,
      "step": 1152
    },
    {
      "epoch": 17.73846153846154,
      "grad_norm": 1.2421875,
      "learning_rate": 2.372726360983723e-05,
      "loss": 0.0106,
      "step": 1153
    },
    {
      "epoch": 17.753846153846155,
      "grad_norm": 3.0,
      "learning_rate": 2.372342441640396e-05,
      "loss": 0.0393,
      "step": 1154
    },
    {
      "epoch": 17.76923076923077,
      "grad_norm": 0.482421875,
      "learning_rate": 2.371957975280997e-05,
      "loss": 0.0086,
      "step": 1155
    },
    {
      "epoch": 17.784615384615385,
      "grad_norm": 0.9296875,
      "learning_rate": 2.3715729620929106e-05,
      "loss": 0.0128,
      "step": 1156
    },
    {
      "epoch": 17.8,
      "grad_norm": 4.09375,
      "learning_rate": 2.3711874022637875e-05,
      "loss": 0.0441,
      "step": 1157
    },
    {
      "epoch": 17.815384615384616,
      "grad_norm": 3.0625,
      "learning_rate": 2.3708012959815452e-05,
      "loss": 0.0668,
      "step": 1158
    },
    {
      "epoch": 17.83076923076923,
      "grad_norm": 3.140625,
      "learning_rate": 2.3704146434343664e-05,
      "loss": 0.1019,
      "step": 1159
    },
    {
      "epoch": 17.846153846153847,
      "grad_norm": 1.4296875,
      "learning_rate": 2.3700274448107012e-05,
      "loss": 0.0104,
      "step": 1160
    },
    {
      "epoch": 17.861538461538462,
      "grad_norm": 3.609375,
      "learning_rate": 2.3696397002992663e-05,
      "loss": 0.0285,
      "step": 1161
    },
    {
      "epoch": 17.876923076923077,
      "grad_norm": 4.625,
      "learning_rate": 2.3692514100890425e-05,
      "loss": 0.0278,
      "step": 1162
    },
    {
      "epoch": 17.892307692307693,
      "grad_norm": 4.46875,
      "learning_rate": 2.3688625743692784e-05,
      "loss": 0.0978,
      "step": 1163
    },
    {
      "epoch": 17.907692307692308,
      "grad_norm": 4.5,
      "learning_rate": 2.368473193329488e-05,
      "loss": 0.0328,
      "step": 1164
    },
    {
      "epoch": 17.923076923076923,
      "grad_norm": 2.421875,
      "learning_rate": 2.36808326715945e-05,
      "loss": 0.0139,
      "step": 1165
    },
    {
      "epoch": 17.93846153846154,
      "grad_norm": 2.625,
      "learning_rate": 2.36769279604921e-05,
      "loss": 0.0428,
      "step": 1166
    },
    {
      "epoch": 17.953846153846154,
      "grad_norm": 3.765625,
      "learning_rate": 2.36730178018908e-05,
      "loss": 0.0476,
      "step": 1167
    },
    {
      "epoch": 17.96923076923077,
      "grad_norm": 0.37890625,
      "learning_rate": 2.3669102197696347e-05,
      "loss": 0.0309,
      "step": 1168
    },
    {
      "epoch": 17.984615384615385,
      "grad_norm": 4.3125,
      "learning_rate": 2.3665181149817173e-05,
      "loss": 0.0496,
      "step": 1169
    },
    {
      "epoch": 18.0,
      "grad_norm": 5.28125,
      "learning_rate": 2.3661254660164337e-05,
      "loss": 0.0666,
      "step": 1170
    },
    {
      "epoch": 18.015384615384615,
      "grad_norm": 3.671875,
      "learning_rate": 2.365732273065157e-05,
      "loss": 0.0361,
      "step": 1171
    },
    {
      "epoch": 18.03076923076923,
      "grad_norm": 1.203125,
      "learning_rate": 2.3653385363195248e-05,
      "loss": 0.0522,
      "step": 1172
    },
    {
      "epoch": 18.046153846153846,
      "grad_norm": 3.3125,
      "learning_rate": 2.3649442559714387e-05,
      "loss": 0.0148,
      "step": 1173
    },
    {
      "epoch": 18.06153846153846,
      "grad_norm": 5.375,
      "learning_rate": 2.3645494322130668e-05,
      "loss": 0.0345,
      "step": 1174
    },
    {
      "epoch": 18.076923076923077,
      "grad_norm": 1.296875,
      "learning_rate": 2.3641540652368415e-05,
      "loss": 0.0709,
      "step": 1175
    },
    {
      "epoch": 18.092307692307692,
      "grad_norm": 3.921875,
      "learning_rate": 2.36375815523546e-05,
      "loss": 0.09,
      "step": 1176
    },
    {
      "epoch": 18.107692307692307,
      "grad_norm": 1.78125,
      "learning_rate": 2.3633617024018832e-05,
      "loss": 0.0272,
      "step": 1177
    },
    {
      "epoch": 18.123076923076923,
      "grad_norm": 3.640625,
      "learning_rate": 2.3629647069293384e-05,
      "loss": 0.0736,
      "step": 1178
    },
    {
      "epoch": 18.138461538461538,
      "grad_norm": 3.5,
      "learning_rate": 2.3625671690113155e-05,
      "loss": 0.0535,
      "step": 1179
    },
    {
      "epoch": 18.153846153846153,
      "grad_norm": 2.5625,
      "learning_rate": 2.3621690888415705e-05,
      "loss": 0.0439,
      "step": 1180
    },
    {
      "epoch": 18.16923076923077,
      "grad_norm": 10.5625,
      "learning_rate": 2.361770466614122e-05,
      "loss": 0.0761,
      "step": 1181
    },
    {
      "epoch": 18.184615384615384,
      "grad_norm": 2.921875,
      "learning_rate": 2.3613713025232542e-05,
      "loss": 0.0551,
      "step": 1182
    },
    {
      "epoch": 18.2,
      "grad_norm": 2.9375,
      "learning_rate": 2.3609715967635145e-05,
      "loss": 0.0352,
      "step": 1183
    },
    {
      "epoch": 18.215384615384615,
      "grad_norm": 3.703125,
      "learning_rate": 2.3605713495297148e-05,
      "loss": 0.0594,
      "step": 1184
    },
    {
      "epoch": 18.23076923076923,
      "grad_norm": 3.328125,
      "learning_rate": 2.360170561016931e-05,
      "loss": 0.0495,
      "step": 1185
    },
    {
      "epoch": 18.246153846153845,
      "grad_norm": 2.21875,
      "learning_rate": 2.3597692314205016e-05,
      "loss": 0.0311,
      "step": 1186
    },
    {
      "epoch": 18.26153846153846,
      "grad_norm": 2.234375,
      "learning_rate": 2.359367360936031e-05,
      "loss": 0.0995,
      "step": 1187
    },
    {
      "epoch": 18.276923076923076,
      "grad_norm": 2.75,
      "learning_rate": 2.358964949759385e-05,
      "loss": 0.0501,
      "step": 1188
    },
    {
      "epoch": 18.29230769230769,
      "grad_norm": 2.078125,
      "learning_rate": 2.358561998086695e-05,
      "loss": 0.0245,
      "step": 1189
    },
    {
      "epoch": 18.307692307692307,
      "grad_norm": 2.9375,
      "learning_rate": 2.3581585061143534e-05,
      "loss": 0.0665,
      "step": 1190
    },
    {
      "epoch": 18.323076923076922,
      "grad_norm": 1.0078125,
      "learning_rate": 2.3577544740390184e-05,
      "loss": 0.0494,
      "step": 1191
    },
    {
      "epoch": 18.338461538461537,
      "grad_norm": 2.078125,
      "learning_rate": 2.3573499020576102e-05,
      "loss": 0.0208,
      "step": 1192
    },
    {
      "epoch": 18.353846153846153,
      "grad_norm": 7.65625,
      "learning_rate": 2.3569447903673116e-05,
      "loss": 0.0729,
      "step": 1193
    },
    {
      "epoch": 18.369230769230768,
      "grad_norm": 2.84375,
      "learning_rate": 2.356539139165569e-05,
      "loss": 0.047,
      "step": 1194
    },
    {
      "epoch": 18.384615384615383,
      "grad_norm": 2.046875,
      "learning_rate": 2.356132948650093e-05,
      "loss": 0.023,
      "step": 1195
    },
    {
      "epoch": 18.4,
      "grad_norm": 2.703125,
      "learning_rate": 2.355726219018855e-05,
      "loss": 0.0716,
      "step": 1196
    },
    {
      "epoch": 18.415384615384614,
      "grad_norm": 1.6796875,
      "learning_rate": 2.3553189504700905e-05,
      "loss": 0.0503,
      "step": 1197
    },
    {
      "epoch": 18.43076923076923,
      "grad_norm": 2.0625,
      "learning_rate": 2.3549111432022973e-05,
      "loss": 0.0312,
      "step": 1198
    },
    {
      "epoch": 18.446153846153845,
      "grad_norm": 2.375,
      "learning_rate": 2.3545027974142353e-05,
      "loss": 0.053,
      "step": 1199
    },
    {
      "epoch": 18.46153846153846,
      "grad_norm": 3.90625,
      "learning_rate": 2.3540939133049273e-05,
      "loss": 0.0645,
      "step": 1200
    },
    {
      "epoch": 18.476923076923075,
      "grad_norm": 3.203125,
      "learning_rate": 2.353684491073659e-05,
      "loss": 0.0289,
      "step": 1201
    },
    {
      "epoch": 18.49230769230769,
      "grad_norm": 3.890625,
      "learning_rate": 2.353274530919977e-05,
      "loss": 0.0588,
      "step": 1202
    },
    {
      "epoch": 18.50769230769231,
      "grad_norm": 3.109375,
      "learning_rate": 2.3528640330436917e-05,
      "loss": 0.0278,
      "step": 1203
    },
    {
      "epoch": 18.523076923076925,
      "grad_norm": 1.9921875,
      "learning_rate": 2.3524529976448744e-05,
      "loss": 0.0163,
      "step": 1204
    },
    {
      "epoch": 18.53846153846154,
      "grad_norm": 1.3828125,
      "learning_rate": 2.352041424923859e-05,
      "loss": 0.0134,
      "step": 1205
    },
    {
      "epoch": 18.553846153846155,
      "grad_norm": 3.015625,
      "learning_rate": 2.35162931508124e-05,
      "loss": 0.038,
      "step": 1206
    },
    {
      "epoch": 18.56923076923077,
      "grad_norm": 5.125,
      "learning_rate": 2.3512166683178766e-05,
      "loss": 0.0419,
      "step": 1207
    },
    {
      "epoch": 18.584615384615386,
      "grad_norm": 0.65234375,
      "learning_rate": 2.3508034848348866e-05,
      "loss": 0.0078,
      "step": 1208
    },
    {
      "epoch": 18.6,
      "grad_norm": 0.498046875,
      "learning_rate": 2.3503897648336503e-05,
      "loss": 0.0108,
      "step": 1209
    },
    {
      "epoch": 18.615384615384617,
      "grad_norm": 2.171875,
      "learning_rate": 2.3499755085158106e-05,
      "loss": 0.0175,
      "step": 1210
    },
    {
      "epoch": 18.630769230769232,
      "grad_norm": 2.0,
      "learning_rate": 2.3495607160832707e-05,
      "loss": 0.0224,
      "step": 1211
    },
    {
      "epoch": 18.646153846153847,
      "grad_norm": 0.8203125,
      "learning_rate": 2.3491453877381953e-05,
      "loss": 0.028,
      "step": 1212
    },
    {
      "epoch": 18.661538461538463,
      "grad_norm": 6.375,
      "learning_rate": 2.3487295236830108e-05,
      "loss": 0.0678,
      "step": 1213
    },
    {
      "epoch": 18.676923076923078,
      "grad_norm": 5.90625,
      "learning_rate": 2.3483131241204037e-05,
      "loss": 0.0342,
      "step": 1214
    },
    {
      "epoch": 18.692307692307693,
      "grad_norm": 0.2041015625,
      "learning_rate": 2.3478961892533223e-05,
      "loss": 0.0101,
      "step": 1215
    },
    {
      "epoch": 18.70769230769231,
      "grad_norm": 2.5625,
      "learning_rate": 2.3474787192849756e-05,
      "loss": 0.0114,
      "step": 1216
    },
    {
      "epoch": 18.723076923076924,
      "grad_norm": 1.4609375,
      "learning_rate": 2.347060714418834e-05,
      "loss": 0.0184,
      "step": 1217
    },
    {
      "epoch": 18.73846153846154,
      "grad_norm": 0.171875,
      "learning_rate": 2.346642174858627e-05,
      "loss": 0.0359,
      "step": 1218
    },
    {
      "epoch": 18.753846153846155,
      "grad_norm": 4.5625,
      "learning_rate": 2.346223100808346e-05,
      "loss": 0.0445,
      "step": 1219
    },
    {
      "epoch": 18.76923076923077,
      "grad_norm": 4.5625,
      "learning_rate": 2.345803492472243e-05,
      "loss": 0.0553,
      "step": 1220
    },
    {
      "epoch": 18.784615384615385,
      "grad_norm": 2.171875,
      "learning_rate": 2.3453833500548295e-05,
      "loss": 0.0462,
      "step": 1221
    },
    {
      "epoch": 18.8,
      "grad_norm": 2.09375,
      "learning_rate": 2.3449626737608784e-05,
      "loss": 0.0161,
      "step": 1222
    },
    {
      "epoch": 18.815384615384616,
      "grad_norm": 0.64453125,
      "learning_rate": 2.3445414637954217e-05,
      "loss": 0.0107,
      "step": 1223
    },
    {
      "epoch": 18.83076923076923,
      "grad_norm": 2.421875,
      "learning_rate": 2.3441197203637522e-05,
      "loss": 0.0334,
      "step": 1224
    },
    {
      "epoch": 18.846153846153847,
      "grad_norm": 4.09375,
      "learning_rate": 2.3436974436714224e-05,
      "loss": 0.0136,
      "step": 1225
    },
    {
      "epoch": 18.861538461538462,
      "grad_norm": 1.2578125,
      "learning_rate": 2.3432746339242448e-05,
      "loss": 0.0231,
      "step": 1226
    },
    {
      "epoch": 18.876923076923077,
      "grad_norm": 5.90625,
      "learning_rate": 2.342851291328292e-05,
      "loss": 0.0512,
      "step": 1227
    },
    {
      "epoch": 18.892307692307693,
      "grad_norm": 5.15625,
      "learning_rate": 2.342427416089896e-05,
      "loss": 0.0199,
      "step": 1228
    },
    {
      "epoch": 18.907692307692308,
      "grad_norm": 3.984375,
      "learning_rate": 2.3420030084156486e-05,
      "loss": 0.0235,
      "step": 1229
    },
    {
      "epoch": 18.923076923076923,
      "grad_norm": 0.56640625,
      "learning_rate": 2.3415780685124004e-05,
      "loss": 0.019,
      "step": 1230
    },
    {
      "epoch": 18.93846153846154,
      "grad_norm": 5.75,
      "learning_rate": 2.341152596587262e-05,
      "loss": 0.0233,
      "step": 1231
    },
    {
      "epoch": 18.953846153846154,
      "grad_norm": 4.28125,
      "learning_rate": 2.3407265928476042e-05,
      "loss": 0.0461,
      "step": 1232
    },
    {
      "epoch": 18.96923076923077,
      "grad_norm": 5.75,
      "learning_rate": 2.340300057501055e-05,
      "loss": 0.03,
      "step": 1233
    },
    {
      "epoch": 18.984615384615385,
      "grad_norm": 2.828125,
      "learning_rate": 2.3398729907555026e-05,
      "loss": 0.0108,
      "step": 1234
    },
    {
      "epoch": 19.0,
      "grad_norm": 3.453125,
      "learning_rate": 2.3394453928190947e-05,
      "loss": 0.0164,
      "step": 1235
    },
    {
      "epoch": 19.015384615384615,
      "grad_norm": 1.359375,
      "learning_rate": 2.339017263900237e-05,
      "loss": 0.0226,
      "step": 1236
    },
    {
      "epoch": 19.03076923076923,
      "grad_norm": 13.1875,
      "learning_rate": 2.3385886042075942e-05,
      "loss": 0.0599,
      "step": 1237
    },
    {
      "epoch": 19.046153846153846,
      "grad_norm": 1.5390625,
      "learning_rate": 2.3381594139500902e-05,
      "loss": 0.015,
      "step": 1238
    },
    {
      "epoch": 19.06153846153846,
      "grad_norm": 3.09375,
      "learning_rate": 2.3377296933369074e-05,
      "loss": 0.0361,
      "step": 1239
    },
    {
      "epoch": 19.076923076923077,
      "grad_norm": 1.03125,
      "learning_rate": 2.337299442577485e-05,
      "loss": 0.0101,
      "step": 1240
    },
    {
      "epoch": 19.092307692307692,
      "grad_norm": 1.84375,
      "learning_rate": 2.3368686618815238e-05,
      "loss": 0.0525,
      "step": 1241
    },
    {
      "epoch": 19.107692307692307,
      "grad_norm": 0.359375,
      "learning_rate": 2.33643735145898e-05,
      "loss": 0.0058,
      "step": 1242
    },
    {
      "epoch": 19.123076923076923,
      "grad_norm": 3.140625,
      "learning_rate": 2.336005511520069e-05,
      "loss": 0.0684,
      "step": 1243
    },
    {
      "epoch": 19.138461538461538,
      "grad_norm": 4.34375,
      "learning_rate": 2.3355731422752652e-05,
      "loss": 0.0428,
      "step": 1244
    },
    {
      "epoch": 19.153846153846153,
      "grad_norm": 8.6875,
      "learning_rate": 2.335140243935299e-05,
      "loss": 0.061,
      "step": 1245
    },
    {
      "epoch": 19.16923076923077,
      "grad_norm": 2.625,
      "learning_rate": 2.334706816711161e-05,
      "loss": 0.0715,
      "step": 1246
    },
    {
      "epoch": 19.184615384615384,
      "grad_norm": 5.34375,
      "learning_rate": 2.3342728608140968e-05,
      "loss": 0.0475,
      "step": 1247
    },
    {
      "epoch": 19.2,
      "grad_norm": 1.953125,
      "learning_rate": 2.3338383764556126e-05,
      "loss": 0.0397,
      "step": 1248
    },
    {
      "epoch": 19.215384615384615,
      "grad_norm": 1.375,
      "learning_rate": 2.3334033638474707e-05,
      "loss": 0.0192,
      "step": 1249
    },
    {
      "epoch": 19.23076923076923,
      "grad_norm": 1.7421875,
      "learning_rate": 2.3329678232016898e-05,
      "loss": 0.0168,
      "step": 1250
    },
    {
      "epoch": 19.246153846153845,
      "grad_norm": 3.75,
      "learning_rate": 2.3325317547305485e-05,
      "loss": 0.0292,
      "step": 1251
    },
    {
      "epoch": 19.26153846153846,
      "grad_norm": 2.671875,
      "learning_rate": 2.3320951586465804e-05,
      "loss": 0.0503,
      "step": 1252
    },
    {
      "epoch": 19.276923076923076,
      "grad_norm": 0.36328125,
      "learning_rate": 2.3316580351625777e-05,
      "loss": 0.1273,
      "step": 1253
    },
    {
      "epoch": 19.29230769230769,
      "grad_norm": 6.125,
      "learning_rate": 2.3312203844915885e-05,
      "loss": 0.0412,
      "step": 1254
    },
    {
      "epoch": 19.307692307692307,
      "grad_norm": 7.34375,
      "learning_rate": 2.3307822068469187e-05,
      "loss": 0.0491,
      "step": 1255
    },
    {
      "epoch": 19.323076923076922,
      "grad_norm": 4.5625,
      "learning_rate": 2.3303435024421312e-05,
      "loss": 0.0551,
      "step": 1256
    },
    {
      "epoch": 19.338461538461537,
      "grad_norm": 14.8125,
      "learning_rate": 2.329904271491044e-05,
      "loss": 0.0635,
      "step": 1257
    },
    {
      "epoch": 19.353846153846153,
      "grad_norm": 1.0390625,
      "learning_rate": 2.329464514207734e-05,
      "loss": 0.0264,
      "step": 1258
    },
    {
      "epoch": 19.369230769230768,
      "grad_norm": 1.4921875,
      "learning_rate": 2.329024230806534e-05,
      "loss": 0.0294,
      "step": 1259
    },
    {
      "epoch": 19.384615384615383,
      "grad_norm": 4.125,
      "learning_rate": 2.3285834215020313e-05,
      "loss": 0.0453,
      "step": 1260
    },
    {
      "epoch": 19.4,
      "grad_norm": 8.125,
      "learning_rate": 2.328142086509072e-05,
      "loss": 0.0494,
      "step": 1261
    },
    {
      "epoch": 19.415384615384614,
      "grad_norm": 2.578125,
      "learning_rate": 2.327700226042757e-05,
      "loss": 0.0304,
      "step": 1262
    },
    {
      "epoch": 19.43076923076923,
      "grad_norm": 1.3359375,
      "learning_rate": 2.327257840318444e-05,
      "loss": 0.0292,
      "step": 1263
    },
    {
      "epoch": 19.446153846153845,
      "grad_norm": 8.125,
      "learning_rate": 2.326814929551746e-05,
      "loss": 0.145,
      "step": 1264
    },
    {
      "epoch": 19.46153846153846,
      "grad_norm": 5.0625,
      "learning_rate": 2.326371493958533e-05,
      "loss": 0.0394,
      "step": 1265
    },
    {
      "epoch": 19.476923076923075,
      "grad_norm": 4.5,
      "learning_rate": 2.32592753375493e-05,
      "loss": 0.089,
      "step": 1266
    },
    {
      "epoch": 19.49230769230769,
      "grad_norm": 5.71875,
      "learning_rate": 2.3254830491573167e-05,
      "loss": 0.0363,
      "step": 1267
    },
    {
      "epoch": 19.50769230769231,
      "grad_norm": 4.9375,
      "learning_rate": 2.325038040382331e-05,
      "loss": 0.025,
      "step": 1268
    },
    {
      "epoch": 19.523076923076925,
      "grad_norm": 6.25,
      "learning_rate": 2.324592507646864e-05,
      "loss": 0.0505,
      "step": 1269
    },
    {
      "epoch": 19.53846153846154,
      "grad_norm": 1.0390625,
      "learning_rate": 2.324146451168063e-05,
      "loss": 0.0161,
      "step": 1270
    },
    {
      "epoch": 19.553846153846155,
      "grad_norm": 3.34375,
      "learning_rate": 2.3236998711633307e-05,
      "loss": 0.0514,
      "step": 1271
    },
    {
      "epoch": 19.56923076923077,
      "grad_norm": 1.2109375,
      "learning_rate": 2.3232527678503246e-05,
      "loss": 0.0232,
      "step": 1272
    },
    {
      "epoch": 19.584615384615386,
      "grad_norm": 3.484375,
      "learning_rate": 2.3228051414469574e-05,
      "loss": 0.0127,
      "step": 1273
    },
    {
      "epoch": 19.6,
      "grad_norm": 2.25,
      "learning_rate": 2.3223569921713966e-05,
      "loss": 0.0107,
      "step": 1274
    },
    {
      "epoch": 19.615384615384617,
      "grad_norm": 7.125,
      "learning_rate": 2.3219083202420654e-05,
      "loss": 0.0555,
      "step": 1275
    },
    {
      "epoch": 19.630769230769232,
      "grad_norm": 3.65625,
      "learning_rate": 2.3214591258776404e-05,
      "loss": 0.0326,
      "step": 1276
    },
    {
      "epoch": 19.646153846153847,
      "grad_norm": 4.375,
      "learning_rate": 2.3210094092970536e-05,
      "loss": 0.0334,
      "step": 1277
    },
    {
      "epoch": 19.661538461538463,
      "grad_norm": 1.1484375,
      "learning_rate": 2.3205591707194918e-05,
      "loss": 0.0131,
      "step": 1278
    },
    {
      "epoch": 19.676923076923078,
      "grad_norm": 4.21875,
      "learning_rate": 2.320108410364395e-05,
      "loss": 0.0279,
      "step": 1279
    },
    {
      "epoch": 19.692307692307693,
      "grad_norm": 2.90625,
      "learning_rate": 2.3196571284514594e-05,
      "loss": 0.0386,
      "step": 1280
    },
    {
      "epoch": 19.70769230769231,
      "grad_norm": 3.734375,
      "learning_rate": 2.3192053252006335e-05,
      "loss": 0.0247,
      "step": 1281
    },
    {
      "epoch": 19.723076923076924,
      "grad_norm": 1.28125,
      "learning_rate": 2.3187530008321208e-05,
      "loss": 0.0101,
      "step": 1282
    },
    {
      "epoch": 19.73846153846154,
      "grad_norm": 6.78125,
      "learning_rate": 2.318300155566379e-05,
      "loss": 0.0329,
      "step": 1283
    },
    {
      "epoch": 19.753846153846155,
      "grad_norm": 9.3125,
      "learning_rate": 2.317846789624119e-05,
      "loss": 0.0322,
      "step": 1284
    },
    {
      "epoch": 19.76923076923077,
      "grad_norm": 5.9375,
      "learning_rate": 2.3173929032263063e-05,
      "loss": 0.0644,
      "step": 1285
    },
    {
      "epoch": 19.784615384615385,
      "grad_norm": 1.1640625,
      "learning_rate": 2.3169384965941592e-05,
      "loss": 0.0262,
      "step": 1286
    },
    {
      "epoch": 19.8,
      "grad_norm": 4.15625,
      "learning_rate": 2.3164835699491505e-05,
      "loss": 0.0207,
      "step": 1287
    },
    {
      "epoch": 19.815384615384616,
      "grad_norm": 3.421875,
      "learning_rate": 2.3160281235130055e-05,
      "loss": 0.051,
      "step": 1288
    },
    {
      "epoch": 19.83076923076923,
      "grad_norm": 2.859375,
      "learning_rate": 2.3155721575077034e-05,
      "loss": 0.0216,
      "step": 1289
    },
    {
      "epoch": 19.846153846153847,
      "grad_norm": 1.40625,
      "learning_rate": 2.3151156721554763e-05,
      "loss": 0.0081,
      "step": 1290
    },
    {
      "epoch": 19.861538461538462,
      "grad_norm": 0.5703125,
      "learning_rate": 2.3146586676788095e-05,
      "loss": 0.0463,
      "step": 1291
    },
    {
      "epoch": 19.876923076923077,
      "grad_norm": 2.8125,
      "learning_rate": 2.3142011443004418e-05,
      "loss": 0.0137,
      "step": 1292
    },
    {
      "epoch": 19.892307692307693,
      "grad_norm": 6.875,
      "learning_rate": 2.3137431022433652e-05,
      "loss": 0.0296,
      "step": 1293
    },
    {
      "epoch": 19.907692307692308,
      "grad_norm": 0.9921875,
      "learning_rate": 2.3132845417308223e-05,
      "loss": 0.0087,
      "step": 1294
    },
    {
      "epoch": 19.923076923076923,
      "grad_norm": 1.2109375,
      "learning_rate": 2.312825462986311e-05,
      "loss": 0.0214,
      "step": 1295
    },
    {
      "epoch": 19.93846153846154,
      "grad_norm": 3.390625,
      "learning_rate": 2.3123658662335802e-05,
      "loss": 0.0152,
      "step": 1296
    },
    {
      "epoch": 19.953846153846154,
      "grad_norm": 1.140625,
      "learning_rate": 2.3119057516966318e-05,
      "loss": 0.0491,
      "step": 1297
    },
    {
      "epoch": 19.96923076923077,
      "grad_norm": 3.296875,
      "learning_rate": 2.3114451195997206e-05,
      "loss": 0.0863,
      "step": 1298
    },
    {
      "epoch": 19.984615384615385,
      "grad_norm": 2.484375,
      "learning_rate": 2.3109839701673525e-05,
      "loss": 0.015,
      "step": 1299
    },
    {
      "epoch": 20.0,
      "grad_norm": 9.375,
      "learning_rate": 2.3105223036242863e-05,
      "loss": 0.0639,
      "step": 1300
    },
    {
      "epoch": 20.015384615384615,
      "grad_norm": 4.09375,
      "learning_rate": 2.3100601201955324e-05,
      "loss": 0.0355,
      "step": 1301
    },
    {
      "epoch": 20.03076923076923,
      "grad_norm": 7.5625,
      "learning_rate": 2.309597420106354e-05,
      "loss": 0.0569,
      "step": 1302
    },
    {
      "epoch": 20.046153846153846,
      "grad_norm": 3.984375,
      "learning_rate": 2.309134203582265e-05,
      "loss": 0.0623,
      "step": 1303
    },
    {
      "epoch": 20.06153846153846,
      "grad_norm": 1.859375,
      "learning_rate": 2.308670470849032e-05,
      "loss": 0.0311,
      "step": 1304
    },
    {
      "epoch": 20.076923076923077,
      "grad_norm": 0.703125,
      "learning_rate": 2.3082062221326724e-05,
      "loss": 0.0217,
      "step": 1305
    },
    {
      "epoch": 20.092307692307692,
      "grad_norm": 2.71875,
      "learning_rate": 2.3077414576594553e-05,
      "loss": 0.0232,
      "step": 1306
    },
    {
      "epoch": 20.107692307692307,
      "grad_norm": 1.4140625,
      "learning_rate": 2.307276177655901e-05,
      "loss": 0.0282,
      "step": 1307
    },
    {
      "epoch": 20.123076923076923,
      "grad_norm": 3.40625,
      "learning_rate": 2.306810382348782e-05,
      "loss": 0.0323,
      "step": 1308
    },
    {
      "epoch": 20.138461538461538,
      "grad_norm": 7.125,
      "learning_rate": 2.306344071965121e-05,
      "loss": 0.0541,
      "step": 1309
    },
    {
      "epoch": 20.153846153846153,
      "grad_norm": 4.71875,
      "learning_rate": 2.3058772467321914e-05,
      "loss": 0.0484,
      "step": 1310
    },
    {
      "epoch": 20.16923076923077,
      "grad_norm": 1.171875,
      "learning_rate": 2.305409906877519e-05,
      "loss": 0.0196,
      "step": 1311
    },
    {
      "epoch": 20.184615384615384,
      "grad_norm": 4.5625,
      "learning_rate": 2.3049420526288792e-05,
      "loss": 0.0358,
      "step": 1312
    },
    {
      "epoch": 20.2,
      "grad_norm": 3.765625,
      "learning_rate": 2.3044736842142987e-05,
      "loss": 0.0254,
      "step": 1313
    },
    {
      "epoch": 20.215384615384615,
      "grad_norm": 5.0625,
      "learning_rate": 2.3040048018620543e-05,
      "loss": 0.033,
      "step": 1314
    },
    {
      "epoch": 20.23076923076923,
      "grad_norm": 7.03125,
      "learning_rate": 2.3035354058006738e-05,
      "loss": 0.0764,
      "step": 1315
    },
    {
      "epoch": 20.246153846153845,
      "grad_norm": 3.09375,
      "learning_rate": 2.3030654962589346e-05,
      "loss": 0.0806,
      "step": 1316
    },
    {
      "epoch": 20.26153846153846,
      "grad_norm": 2.34375,
      "learning_rate": 2.3025950734658654e-05,
      "loss": 0.0461,
      "step": 1317
    },
    {
      "epoch": 20.276923076923076,
      "grad_norm": 2.5,
      "learning_rate": 2.3021241376507447e-05,
      "loss": 0.0161,
      "step": 1318
    },
    {
      "epoch": 20.29230769230769,
      "grad_norm": 5.625,
      "learning_rate": 2.3016526890431006e-05,
      "loss": 0.0566,
      "step": 1319
    },
    {
      "epoch": 20.307692307692307,
      "grad_norm": 9.9375,
      "learning_rate": 2.3011807278727116e-05,
      "loss": 0.076,
      "step": 1320
    },
    {
      "epoch": 20.323076923076922,
      "grad_norm": 1.9765625,
      "learning_rate": 2.3007082543696055e-05,
      "loss": 0.0179,
      "step": 1321
    },
    {
      "epoch": 20.338461538461537,
      "grad_norm": 5.03125,
      "learning_rate": 2.3002352687640605e-05,
      "loss": 0.0463,
      "step": 1322
    },
    {
      "epoch": 20.353846153846153,
      "grad_norm": 4.25,
      "learning_rate": 2.2997617712866037e-05,
      "loss": 0.0405,
      "step": 1323
    },
    {
      "epoch": 20.369230769230768,
      "grad_norm": 2.65625,
      "learning_rate": 2.2992877621680125e-05,
      "loss": 0.0425,
      "step": 1324
    },
    {
      "epoch": 20.384615384615383,
      "grad_norm": 3.15625,
      "learning_rate": 2.2988132416393125e-05,
      "loss": 0.0369,
      "step": 1325
    },
    {
      "epoch": 20.4,
      "grad_norm": 2.171875,
      "learning_rate": 2.2983382099317803e-05,
      "loss": 0.0189,
      "step": 1326
    },
    {
      "epoch": 20.415384615384614,
      "grad_norm": 5.03125,
      "learning_rate": 2.2978626672769395e-05,
      "loss": 0.0468,
      "step": 1327
    },
    {
      "epoch": 20.43076923076923,
      "grad_norm": 4.40625,
      "learning_rate": 2.2973866139065643e-05,
      "loss": 0.0412,
      "step": 1328
    },
    {
      "epoch": 20.446153846153845,
      "grad_norm": 3.359375,
      "learning_rate": 2.2969100500526775e-05,
      "loss": 0.0774,
      "step": 1329
    },
    {
      "epoch": 20.46153846153846,
      "grad_norm": 2.609375,
      "learning_rate": 2.29643297594755e-05,
      "loss": 0.027,
      "step": 1330
    },
    {
      "epoch": 20.476923076923075,
      "grad_norm": 0.5625,
      "learning_rate": 2.295955391823702e-05,
      "loss": 0.0928,
      "step": 1331
    },
    {
      "epoch": 20.49230769230769,
      "grad_norm": 4.4375,
      "learning_rate": 2.295477297913903e-05,
      "loss": 0.0466,
      "step": 1332
    },
    {
      "epoch": 20.50769230769231,
      "grad_norm": 7.125,
      "learning_rate": 2.294998694451169e-05,
      "loss": 0.0566,
      "step": 1333
    },
    {
      "epoch": 20.523076923076925,
      "grad_norm": 1.3515625,
      "learning_rate": 2.294519581668767e-05,
      "loss": 0.0192,
      "step": 1334
    },
    {
      "epoch": 20.53846153846154,
      "grad_norm": 6.125,
      "learning_rate": 2.2940399598002095e-05,
      "loss": 0.0641,
      "step": 1335
    },
    {
      "epoch": 20.553846153846155,
      "grad_norm": 2.796875,
      "learning_rate": 2.2935598290792583e-05,
      "loss": 0.0214,
      "step": 1336
    },
    {
      "epoch": 20.56923076923077,
      "grad_norm": 1.546875,
      "learning_rate": 2.293079189739924e-05,
      "loss": 0.0302,
      "step": 1337
    },
    {
      "epoch": 20.584615384615386,
      "grad_norm": 3.78125,
      "learning_rate": 2.292598042016464e-05,
      "loss": 0.0298,
      "step": 1338
    },
    {
      "epoch": 20.6,
      "grad_norm": 0.7265625,
      "learning_rate": 2.292116386143384e-05,
      "loss": 0.0276,
      "step": 1339
    },
    {
      "epoch": 20.615384615384617,
      "grad_norm": 2.875,
      "learning_rate": 2.2916342223554374e-05,
      "loss": 0.0402,
      "step": 1340
    },
    {
      "epoch": 20.630769230769232,
      "grad_norm": 0.87890625,
      "learning_rate": 2.2911515508876243e-05,
      "loss": 0.0138,
      "step": 1341
    },
    {
      "epoch": 20.646153846153847,
      "grad_norm": 1.828125,
      "learning_rate": 2.2906683719751938e-05,
      "loss": 0.0223,
      "step": 1342
    },
    {
      "epoch": 20.661538461538463,
      "grad_norm": 3.5,
      "learning_rate": 2.2901846858536408e-05,
      "loss": 0.0449,
      "step": 1343
    },
    {
      "epoch": 20.676923076923078,
      "grad_norm": 2.578125,
      "learning_rate": 2.289700492758708e-05,
      "loss": 0.0361,
      "step": 1344
    },
    {
      "epoch": 20.692307692307693,
      "grad_norm": 1.921875,
      "learning_rate": 2.2892157929263858e-05,
      "loss": 0.0096,
      "step": 1345
    },
    {
      "epoch": 20.70769230769231,
      "grad_norm": 4.40625,
      "learning_rate": 2.2887305865929104e-05,
      "loss": 0.0298,
      "step": 1346
    },
    {
      "epoch": 20.723076923076924,
      "grad_norm": 6.25,
      "learning_rate": 2.2882448739947658e-05,
      "loss": 0.0734,
      "step": 1347
    },
    {
      "epoch": 20.73846153846154,
      "grad_norm": 3.0625,
      "learning_rate": 2.2877586553686824e-05,
      "loss": 0.0427,
      "step": 1348
    },
    {
      "epoch": 20.753846153846155,
      "grad_norm": 0.82421875,
      "learning_rate": 2.287271930951637e-05,
      "loss": 0.0195,
      "step": 1349
    },
    {
      "epoch": 20.76923076923077,
      "grad_norm": 2.3125,
      "learning_rate": 2.286784700980854e-05,
      "loss": 0.0262,
      "step": 1350
    },
    {
      "epoch": 20.784615384615385,
      "grad_norm": 1.984375,
      "learning_rate": 2.286296965693802e-05,
      "loss": 0.0154,
      "step": 1351
    },
    {
      "epoch": 20.8,
      "grad_norm": 3.015625,
      "learning_rate": 2.2858087253281986e-05,
      "loss": 0.0305,
      "step": 1352
    },
    {
      "epoch": 20.815384615384616,
      "grad_norm": 0.91015625,
      "learning_rate": 2.2853199801220053e-05,
      "loss": 0.0095,
      "step": 1353
    },
    {
      "epoch": 20.83076923076923,
      "grad_norm": 4.46875,
      "learning_rate": 2.284830730313431e-05,
      "loss": 0.0487,
      "step": 1354
    },
    {
      "epoch": 20.846153846153847,
      "grad_norm": 0.48828125,
      "learning_rate": 2.2843409761409297e-05,
      "loss": 0.0188,
      "step": 1355
    },
    {
      "epoch": 20.861538461538462,
      "grad_norm": 3.703125,
      "learning_rate": 2.283850717843202e-05,
      "loss": 0.0212,
      "step": 1356
    },
    {
      "epoch": 20.876923076923077,
      "grad_norm": 4.09375,
      "learning_rate": 2.2833599556591945e-05,
      "loss": 0.0425,
      "step": 1357
    },
    {
      "epoch": 20.892307692307693,
      "grad_norm": 1.65625,
      "learning_rate": 2.2828686898280974e-05,
      "loss": 0.0074,
      "step": 1358
    },
    {
      "epoch": 20.907692307692308,
      "grad_norm": 4.3125,
      "learning_rate": 2.2823769205893484e-05,
      "loss": 0.0314,
      "step": 1359
    },
    {
      "epoch": 20.923076923076923,
      "grad_norm": 4.90625,
      "learning_rate": 2.2818846481826298e-05,
      "loss": 0.0298,
      "step": 1360
    },
    {
      "epoch": 20.93846153846154,
      "grad_norm": 6.6875,
      "learning_rate": 2.281391872847869e-05,
      "loss": 0.0719,
      "step": 1361
    },
    {
      "epoch": 20.953846153846154,
      "grad_norm": 4.28125,
      "learning_rate": 2.280898594825239e-05,
      "loss": 0.0207,
      "step": 1362
    },
    {
      "epoch": 20.96923076923077,
      "grad_norm": 6.9375,
      "learning_rate": 2.2804048143551575e-05,
      "loss": 0.0291,
      "step": 1363
    },
    {
      "epoch": 20.984615384615385,
      "grad_norm": 1.6796875,
      "learning_rate": 2.279910531678287e-05,
      "loss": 0.0197,
      "step": 1364
    },
    {
      "epoch": 21.0,
      "grad_norm": 2.15625,
      "learning_rate": 2.2794157470355344e-05,
      "loss": 0.0617,
      "step": 1365
    },
    {
      "epoch": 21.015384615384615,
      "grad_norm": 0.52734375,
      "learning_rate": 2.2789204606680524e-05,
      "loss": 0.0058,
      "step": 1366
    },
    {
      "epoch": 21.03076923076923,
      "grad_norm": 0.5625,
      "learning_rate": 2.2784246728172375e-05,
      "loss": 0.0079,
      "step": 1367
    },
    {
      "epoch": 21.046153846153846,
      "grad_norm": 0.46484375,
      "learning_rate": 2.2779283837247305e-05,
      "loss": 0.0056,
      "step": 1368
    },
    {
      "epoch": 21.06153846153846,
      "grad_norm": 1.9375,
      "learning_rate": 2.2774315936324163e-05,
      "loss": 0.0639,
      "step": 1369
    },
    {
      "epoch": 21.076923076923077,
      "grad_norm": 0.67578125,
      "learning_rate": 2.2769343027824253e-05,
      "loss": 0.0199,
      "step": 1370
    },
    {
      "epoch": 21.092307692307692,
      "grad_norm": 3.984375,
      "learning_rate": 2.2764365114171303e-05,
      "loss": 0.0581,
      "step": 1371
    },
    {
      "epoch": 21.107692307692307,
      "grad_norm": 4.5625,
      "learning_rate": 2.275938219779149e-05,
      "loss": 0.0195,
      "step": 1372
    },
    {
      "epoch": 21.123076923076923,
      "grad_norm": 0.2021484375,
      "learning_rate": 2.2754394281113424e-05,
      "loss": 0.0212,
      "step": 1373
    },
    {
      "epoch": 21.138461538461538,
      "grad_norm": 7.90625,
      "learning_rate": 2.2749401366568154e-05,
      "loss": 0.0442,
      "step": 1374
    },
    {
      "epoch": 21.153846153846153,
      "grad_norm": 3.171875,
      "learning_rate": 2.2744403456589173e-05,
      "loss": 0.054,
      "step": 1375
    },
    {
      "epoch": 21.16923076923077,
      "grad_norm": 0.51953125,
      "learning_rate": 2.27394005536124e-05,
      "loss": 0.0083,
      "step": 1376
    },
    {
      "epoch": 21.184615384615384,
      "grad_norm": 2.140625,
      "learning_rate": 2.2734392660076183e-05,
      "loss": 0.0223,
      "step": 1377
    },
    {
      "epoch": 21.2,
      "grad_norm": 3.3125,
      "learning_rate": 2.2729379778421312e-05,
      "loss": 0.0329,
      "step": 1378
    },
    {
      "epoch": 21.215384615384615,
      "grad_norm": 1.140625,
      "learning_rate": 2.272436191109101e-05,
      "loss": 0.018,
      "step": 1379
    },
    {
      "epoch": 21.23076923076923,
      "grad_norm": 8.1875,
      "learning_rate": 2.2719339060530914e-05,
      "loss": 0.0427,
      "step": 1380
    },
    {
      "epoch": 21.246153846153845,
      "grad_norm": 0.96875,
      "learning_rate": 2.2714311229189105e-05,
      "loss": 0.0135,
      "step": 1381
    },
    {
      "epoch": 21.26153846153846,
      "grad_norm": 0.828125,
      "learning_rate": 2.2709278419516096e-05,
      "loss": 0.0042,
      "step": 1382
    },
    {
      "epoch": 21.276923076923076,
      "grad_norm": 2.921875,
      "learning_rate": 2.2704240633964802e-05,
      "loss": 0.027,
      "step": 1383
    },
    {
      "epoch": 21.29230769230769,
      "grad_norm": 3.890625,
      "learning_rate": 2.2699197874990584e-05,
      "loss": 0.0336,
      "step": 1384
    },
    {
      "epoch": 21.307692307692307,
      "grad_norm": 1.3046875,
      "learning_rate": 2.2694150145051228e-05,
      "loss": 0.011,
      "step": 1385
    },
    {
      "epoch": 21.323076923076922,
      "grad_norm": 4.71875,
      "learning_rate": 2.268909744660693e-05,
      "loss": 0.034,
      "step": 1386
    },
    {
      "epoch": 21.338461538461537,
      "grad_norm": 1.6171875,
      "learning_rate": 2.2684039782120315e-05,
      "loss": 0.0576,
      "step": 1387
    },
    {
      "epoch": 21.353846153846153,
      "grad_norm": 2.3125,
      "learning_rate": 2.2678977154056426e-05,
      "loss": 0.0229,
      "step": 1388
    },
    {
      "epoch": 21.369230769230768,
      "grad_norm": 2.671875,
      "learning_rate": 2.267390956488273e-05,
      "loss": 0.0809,
      "step": 1389
    },
    {
      "epoch": 21.384615384615383,
      "grad_norm": 6.78125,
      "learning_rate": 2.26688370170691e-05,
      "loss": 0.0223,
      "step": 1390
    },
    {
      "epoch": 21.4,
      "grad_norm": 5.4375,
      "learning_rate": 2.2663759513087846e-05,
      "loss": 0.0349,
      "step": 1391
    },
    {
      "epoch": 21.415384615384614,
      "grad_norm": 6.6875,
      "learning_rate": 2.2658677055413676e-05,
      "loss": 0.0425,
      "step": 1392
    },
    {
      "epoch": 21.43076923076923,
      "grad_norm": 0.63671875,
      "learning_rate": 2.2653589646523714e-05,
      "loss": 0.0387,
      "step": 1393
    },
    {
      "epoch": 21.446153846153845,
      "grad_norm": 1.1796875,
      "learning_rate": 2.2648497288897514e-05,
      "loss": 0.0281,
      "step": 1394
    },
    {
      "epoch": 21.46153846153846,
      "grad_norm": 0.90234375,
      "learning_rate": 2.2643399985017015e-05,
      "loss": 0.0089,
      "step": 1395
    },
    {
      "epoch": 21.476923076923075,
      "grad_norm": 5.0,
      "learning_rate": 2.2638297737366583e-05,
      "loss": 0.0269,
      "step": 1396
    },
    {
      "epoch": 21.49230769230769,
      "grad_norm": 2.703125,
      "learning_rate": 2.2633190548433e-05,
      "loss": 0.03,
      "step": 1397
    },
    {
      "epoch": 21.50769230769231,
      "grad_norm": 5.625,
      "learning_rate": 2.2628078420705443e-05,
      "loss": 0.0184,
      "step": 1398
    },
    {
      "epoch": 21.523076923076925,
      "grad_norm": 4.53125,
      "learning_rate": 2.2622961356675498e-05,
      "loss": 0.045,
      "step": 1399
    },
    {
      "epoch": 21.53846153846154,
      "grad_norm": 1.015625,
      "learning_rate": 2.261783935883717e-05,
      "loss": 0.0226,
      "step": 1400
    },
    {
      "epoch": 21.553846153846155,
      "grad_norm": 0.482421875,
      "learning_rate": 2.2612712429686845e-05,
      "loss": 0.005,
      "step": 1401
    },
    {
      "epoch": 21.56923076923077,
      "grad_norm": 5.75,
      "learning_rate": 2.260758057172333e-05,
      "loss": 0.059,
      "step": 1402
    },
    {
      "epoch": 21.584615384615386,
      "grad_norm": 1.265625,
      "learning_rate": 2.2602443787447842e-05,
      "loss": 0.0117,
      "step": 1403
    },
    {
      "epoch": 21.6,
      "grad_norm": 7.34375,
      "learning_rate": 2.259730207936397e-05,
      "loss": 0.0317,
      "step": 1404
    },
    {
      "epoch": 21.615384615384617,
      "grad_norm": 9.25,
      "learning_rate": 2.2592155449977733e-05,
      "loss": 0.1035,
      "step": 1405
    },
    {
      "epoch": 21.630769230769232,
      "grad_norm": 3.953125,
      "learning_rate": 2.2587003901797528e-05,
      "loss": 0.0376,
      "step": 1406
    },
    {
      "epoch": 21.646153846153847,
      "grad_norm": 4.6875,
      "learning_rate": 2.258184743733416e-05,
      "loss": 0.032,
      "step": 1407
    },
    {
      "epoch": 21.661538461538463,
      "grad_norm": 3.6875,
      "learning_rate": 2.2576686059100825e-05,
      "loss": 0.0205,
      "step": 1408
    },
    {
      "epoch": 21.676923076923078,
      "grad_norm": 10.5,
      "learning_rate": 2.2571519769613117e-05,
      "loss": 0.0656,
      "step": 1409
    },
    {
      "epoch": 21.692307692307693,
      "grad_norm": 2.28125,
      "learning_rate": 2.2566348571389025e-05,
      "loss": 0.0353,
      "step": 1410
    },
    {
      "epoch": 21.70769230769231,
      "grad_norm": 2.484375,
      "learning_rate": 2.2561172466948925e-05,
      "loss": 0.0313,
      "step": 1411
    },
    {
      "epoch": 21.723076923076924,
      "grad_norm": 1.9140625,
      "learning_rate": 2.2555991458815597e-05,
      "loss": 0.0267,
      "step": 1412
    },
    {
      "epoch": 21.73846153846154,
      "grad_norm": 1.359375,
      "learning_rate": 2.2550805549514184e-05,
      "loss": 0.0222,
      "step": 1413
    },
    {
      "epoch": 21.753846153846155,
      "grad_norm": 1.296875,
      "learning_rate": 2.2545614741572248e-05,
      "loss": 0.0249,
      "step": 1414
    },
    {
      "epoch": 21.76923076923077,
      "grad_norm": 3.46875,
      "learning_rate": 2.2540419037519725e-05,
      "loss": 0.023,
      "step": 1415
    },
    {
      "epoch": 21.784615384615385,
      "grad_norm": 2.484375,
      "learning_rate": 2.2535218439888933e-05,
      "loss": 0.0197,
      "step": 1416
    },
    {
      "epoch": 21.8,
      "grad_norm": 1.765625,
      "learning_rate": 2.2530012951214587e-05,
      "loss": 0.0276,
      "step": 1417
    },
    {
      "epoch": 21.815384615384616,
      "grad_norm": 0.53125,
      "learning_rate": 2.2524802574033765e-05,
      "loss": 0.0047,
      "step": 1418
    },
    {
      "epoch": 21.83076923076923,
      "grad_norm": 2.65625,
      "learning_rate": 2.251958731088596e-05,
      "loss": 0.0451,
      "step": 1419
    },
    {
      "epoch": 21.846153846153847,
      "grad_norm": 4.1875,
      "learning_rate": 2.2514367164313017e-05,
      "loss": 0.0242,
      "step": 1420
    },
    {
      "epoch": 21.861538461538462,
      "grad_norm": 3.359375,
      "learning_rate": 2.2509142136859168e-05,
      "loss": 0.0221,
      "step": 1421
    },
    {
      "epoch": 21.876923076923077,
      "grad_norm": 1.7890625,
      "learning_rate": 2.2503912231071038e-05,
      "loss": 0.0202,
      "step": 1422
    },
    {
      "epoch": 21.892307692307693,
      "grad_norm": 0.306640625,
      "learning_rate": 2.249867744949762e-05,
      "loss": 0.0218,
      "step": 1423
    },
    {
      "epoch": 21.907692307692308,
      "grad_norm": 1.0625,
      "learning_rate": 2.2493437794690274e-05,
      "loss": 0.019,
      "step": 1424
    },
    {
      "epoch": 21.923076923076923,
      "grad_norm": 1.75,
      "learning_rate": 2.2488193269202748e-05,
      "loss": 0.0118,
      "step": 1425
    },
    {
      "epoch": 21.93846153846154,
      "grad_norm": 3.375,
      "learning_rate": 2.248294387559116e-05,
      "loss": 0.0249,
      "step": 1426
    },
    {
      "epoch": 21.953846153846154,
      "grad_norm": 1.375,
      "learning_rate": 2.2477689616414004e-05,
      "loss": 0.0086,
      "step": 1427
    },
    {
      "epoch": 21.96923076923077,
      "grad_norm": 7.78125,
      "learning_rate": 2.2472430494232138e-05,
      "loss": 0.0681,
      "step": 1428
    },
    {
      "epoch": 21.984615384615385,
      "grad_norm": 3.75,
      "learning_rate": 2.2467166511608796e-05,
      "loss": 0.0514,
      "step": 1429
    },
    {
      "epoch": 22.0,
      "grad_norm": 4.0,
      "learning_rate": 2.246189767110958e-05,
      "loss": 0.0201,
      "step": 1430
    },
    {
      "epoch": 22.015384615384615,
      "grad_norm": 6.59375,
      "learning_rate": 2.2456623975302454e-05,
      "loss": 0.0474,
      "step": 1431
    },
    {
      "epoch": 22.03076923076923,
      "grad_norm": 1.140625,
      "learning_rate": 2.245134542675776e-05,
      "loss": 0.0177,
      "step": 1432
    },
    {
      "epoch": 22.046153846153846,
      "grad_norm": 0.439453125,
      "learning_rate": 2.2446062028048193e-05,
      "loss": 0.0193,
      "step": 1433
    },
    {
      "epoch": 22.06153846153846,
      "grad_norm": 3.46875,
      "learning_rate": 2.244077378174882e-05,
      "loss": 0.017,
      "step": 1434
    },
    {
      "epoch": 22.076923076923077,
      "grad_norm": 4.4375,
      "learning_rate": 2.243548069043707e-05,
      "loss": 0.0351,
      "step": 1435
    },
    {
      "epoch": 22.092307692307692,
      "grad_norm": 0.99609375,
      "learning_rate": 2.2430182756692728e-05,
      "loss": 0.0106,
      "step": 1436
    },
    {
      "epoch": 22.107692307692307,
      "grad_norm": 1.625,
      "learning_rate": 2.2424879983097943e-05,
      "loss": 0.0208,
      "step": 1437
    },
    {
      "epoch": 22.123076923076923,
      "grad_norm": 3.375,
      "learning_rate": 2.2419572372237226e-05,
      "loss": 0.0223,
      "step": 1438
    },
    {
      "epoch": 22.138461538461538,
      "grad_norm": 1.625,
      "learning_rate": 2.2414259926697434e-05,
      "loss": 0.0128,
      "step": 1439
    },
    {
      "epoch": 22.153846153846153,
      "grad_norm": 4.625,
      "learning_rate": 2.2408942649067795e-05,
      "loss": 0.0545,
      "step": 1440
    },
    {
      "epoch": 22.16923076923077,
      "grad_norm": 1.1015625,
      "learning_rate": 2.2403620541939884e-05,
      "loss": 0.0088,
      "step": 1441
    },
    {
      "epoch": 22.184615384615384,
      "grad_norm": 3.765625,
      "learning_rate": 2.2398293607907634e-05,
      "loss": 0.0227,
      "step": 1442
    },
    {
      "epoch": 22.2,
      "grad_norm": 7.09375,
      "learning_rate": 2.2392961849567325e-05,
      "loss": 0.0415,
      "step": 1443
    },
    {
      "epoch": 22.215384615384615,
      "grad_norm": 1.1875,
      "learning_rate": 2.2387625269517587e-05,
      "loss": 0.0137,
      "step": 1444
    },
    {
      "epoch": 22.23076923076923,
      "grad_norm": 4.03125,
      "learning_rate": 2.2382283870359412e-05,
      "loss": 0.0458,
      "step": 1445
    },
    {
      "epoch": 22.246153846153845,
      "grad_norm": 1.875,
      "learning_rate": 2.2376937654696133e-05,
      "loss": 0.0501,
      "step": 1446
    },
    {
      "epoch": 22.26153846153846,
      "grad_norm": 0.96484375,
      "learning_rate": 2.237158662513342e-05,
      "loss": 0.045,
      "step": 1447
    },
    {
      "epoch": 22.276923076923076,
      "grad_norm": 7.21875,
      "learning_rate": 2.2366230784279313e-05,
      "loss": 0.0832,
      "step": 1448
    },
    {
      "epoch": 22.29230769230769,
      "grad_norm": 1.1015625,
      "learning_rate": 2.2360870134744178e-05,
      "loss": 0.025,
      "step": 1449
    },
    {
      "epoch": 22.307692307692307,
      "grad_norm": 2.828125,
      "learning_rate": 2.2355504679140728e-05,
      "loss": 0.0477,
      "step": 1450
    },
    {
      "epoch": 22.323076923076922,
      "grad_norm": 8.5,
      "learning_rate": 2.2350134420084023e-05,
      "loss": 0.0739,
      "step": 1451
    },
    {
      "epoch": 22.338461538461537,
      "grad_norm": 0.7734375,
      "learning_rate": 2.2344759360191468e-05,
      "loss": 0.0063,
      "step": 1452
    },
    {
      "epoch": 22.353846153846153,
      "grad_norm": 4.21875,
      "learning_rate": 2.2339379502082792e-05,
      "loss": 0.0655,
      "step": 1453
    },
    {
      "epoch": 22.369230769230768,
      "grad_norm": 2.375,
      "learning_rate": 2.2333994848380075e-05,
      "loss": 0.0203,
      "step": 1454
    },
    {
      "epoch": 22.384615384615383,
      "grad_norm": 2.578125,
      "learning_rate": 2.2328605401707735e-05,
      "loss": 0.0322,
      "step": 1455
    },
    {
      "epoch": 22.4,
      "grad_norm": 8.875,
      "learning_rate": 2.2323211164692526e-05,
      "loss": 0.0682,
      "step": 1456
    },
    {
      "epoch": 22.415384615384614,
      "grad_norm": 0.8203125,
      "learning_rate": 2.2317812139963522e-05,
      "loss": 0.0093,
      "step": 1457
    },
    {
      "epoch": 22.43076923076923,
      "grad_norm": 1.3984375,
      "learning_rate": 2.2312408330152157e-05,
      "loss": 0.0388,
      "step": 1458
    },
    {
      "epoch": 22.446153846153845,
      "grad_norm": 3.765625,
      "learning_rate": 2.2306999737892168e-05,
      "loss": 0.0313,
      "step": 1459
    },
    {
      "epoch": 22.46153846153846,
      "grad_norm": 1.109375,
      "learning_rate": 2.2301586365819644e-05,
      "loss": 0.0168,
      "step": 1460
    },
    {
      "epoch": 22.476923076923075,
      "grad_norm": 4.46875,
      "learning_rate": 2.2296168216573e-05,
      "loss": 0.024,
      "step": 1461
    },
    {
      "epoch": 22.49230769230769,
      "grad_norm": 6.21875,
      "learning_rate": 2.229074529279297e-05,
      "loss": 0.0373,
      "step": 1462
    },
    {
      "epoch": 22.50769230769231,
      "grad_norm": 5.15625,
      "learning_rate": 2.2285317597122623e-05,
      "loss": 0.0524,
      "step": 1463
    },
    {
      "epoch": 22.523076923076925,
      "grad_norm": 1.5078125,
      "learning_rate": 2.227988513220735e-05,
      "loss": 0.0271,
      "step": 1464
    },
    {
      "epoch": 22.53846153846154,
      "grad_norm": 2.96875,
      "learning_rate": 2.2274447900694876e-05,
      "loss": 0.0693,
      "step": 1465
    },
    {
      "epoch": 22.553846153846155,
      "grad_norm": 4.15625,
      "learning_rate": 2.2269005905235234e-05,
      "loss": 0.0304,
      "step": 1466
    },
    {
      "epoch": 22.56923076923077,
      "grad_norm": 2.75,
      "learning_rate": 2.226355914848079e-05,
      "loss": 0.0215,
      "step": 1467
    },
    {
      "epoch": 22.584615384615386,
      "grad_norm": 0.7890625,
      "learning_rate": 2.2258107633086227e-05,
      "loss": 0.0084,
      "step": 1468
    },
    {
      "epoch": 22.6,
      "grad_norm": 3.34375,
      "learning_rate": 2.2252651361708544e-05,
      "loss": 0.0366,
      "step": 1469
    },
    {
      "epoch": 22.615384615384617,
      "grad_norm": 4.78125,
      "learning_rate": 2.2247190337007066e-05,
      "loss": 0.053,
      "step": 1470
    },
    {
      "epoch": 22.630769230769232,
      "grad_norm": 2.921875,
      "learning_rate": 2.224172456164343e-05,
      "loss": 0.0092,
      "step": 1471
    },
    {
      "epoch": 22.646153846153847,
      "grad_norm": 1.890625,
      "learning_rate": 2.2236254038281583e-05,
      "loss": 0.0285,
      "step": 1472
    },
    {
      "epoch": 22.661538461538463,
      "grad_norm": 2.09375,
      "learning_rate": 2.2230778769587797e-05,
      "loss": 0.07,
      "step": 1473
    },
    {
      "epoch": 22.676923076923078,
      "grad_norm": 0.8515625,
      "learning_rate": 2.2225298758230643e-05,
      "loss": 0.0442,
      "step": 1474
    },
    {
      "epoch": 22.692307692307693,
      "grad_norm": 1.8359375,
      "learning_rate": 2.2219814006881027e-05,
      "loss": 0.0506,
      "step": 1475
    },
    {
      "epoch": 22.70769230769231,
      "grad_norm": 3.46875,
      "learning_rate": 2.221432451821214e-05,
      "loss": 0.0215,
      "step": 1476
    },
    {
      "epoch": 22.723076923076924,
      "grad_norm": 6.6875,
      "learning_rate": 2.220883029489949e-05,
      "loss": 0.0325,
      "step": 1477
    },
    {
      "epoch": 22.73846153846154,
      "grad_norm": 4.1875,
      "learning_rate": 2.22033313396209e-05,
      "loss": 0.0283,
      "step": 1478
    },
    {
      "epoch": 22.753846153846155,
      "grad_norm": 4.9375,
      "learning_rate": 2.2197827655056494e-05,
      "loss": 0.0356,
      "step": 1479
    },
    {
      "epoch": 22.76923076923077,
      "grad_norm": 2.125,
      "learning_rate": 2.21923192438887e-05,
      "loss": 0.0302,
      "step": 1480
    },
    {
      "epoch": 22.784615384615385,
      "grad_norm": 10.0625,
      "learning_rate": 2.2186806108802248e-05,
      "loss": 0.0568,
      "step": 1481
    },
    {
      "epoch": 22.8,
      "grad_norm": 1.421875,
      "learning_rate": 2.218128825248417e-05,
      "loss": 0.029,
      "step": 1482
    },
    {
      "epoch": 22.815384615384616,
      "grad_norm": 0.62109375,
      "learning_rate": 2.2175765677623816e-05,
      "loss": 0.0304,
      "step": 1483
    },
    {
      "epoch": 22.83076923076923,
      "grad_norm": 1.484375,
      "learning_rate": 2.2170238386912813e-05,
      "loss": 0.0318,
      "step": 1484
    },
    {
      "epoch": 22.846153846153847,
      "grad_norm": 1.90625,
      "learning_rate": 2.216470638304509e-05,
      "loss": 0.0303,
      "step": 1485
    },
    {
      "epoch": 22.861538461538462,
      "grad_norm": 2.125,
      "learning_rate": 2.215916966871689e-05,
      "loss": 0.0374,
      "step": 1486
    },
    {
      "epoch": 22.876923076923077,
      "grad_norm": 2.578125,
      "learning_rate": 2.2153628246626728e-05,
      "loss": 0.032,
      "step": 1487
    },
    {
      "epoch": 22.892307692307693,
      "grad_norm": 4.40625,
      "learning_rate": 2.2148082119475437e-05,
      "loss": 0.0489,
      "step": 1488
    },
    {
      "epoch": 22.907692307692308,
      "grad_norm": 6.96875,
      "learning_rate": 2.214253128996613e-05,
      "loss": 0.0337,
      "step": 1489
    },
    {
      "epoch": 22.923076923076923,
      "grad_norm": 0.52734375,
      "learning_rate": 2.2136975760804204e-05,
      "loss": 0.0619,
      "step": 1490
    },
    {
      "epoch": 22.93846153846154,
      "grad_norm": 2.8125,
      "learning_rate": 2.213141553469737e-05,
      "loss": 0.0301,
      "step": 1491
    },
    {
      "epoch": 22.953846153846154,
      "grad_norm": 3.359375,
      "learning_rate": 2.21258506143556e-05,
      "loss": 0.0342,
      "step": 1492
    },
    {
      "epoch": 22.96923076923077,
      "grad_norm": 2.125,
      "learning_rate": 2.2120281002491182e-05,
      "loss": 0.0264,
      "step": 1493
    },
    {
      "epoch": 22.984615384615385,
      "grad_norm": 1.71875,
      "learning_rate": 2.2114706701818672e-05,
      "loss": 0.0252,
      "step": 1494
    },
    {
      "epoch": 23.0,
      "grad_norm": 1.53125,
      "learning_rate": 2.2109127715054916e-05,
      "loss": 0.0127,
      "step": 1495
    },
    {
      "epoch": 23.015384615384615,
      "grad_norm": 3.9375,
      "learning_rate": 2.2103544044919045e-05,
      "loss": 0.0377,
      "step": 1496
    },
    {
      "epoch": 23.03076923076923,
      "grad_norm": 4.15625,
      "learning_rate": 2.2097955694132473e-05,
      "loss": 0.0378,
      "step": 1497
    },
    {
      "epoch": 23.046153846153846,
      "grad_norm": 7.03125,
      "learning_rate": 2.2092362665418887e-05,
      "loss": 0.0585,
      "step": 1498
    },
    {
      "epoch": 23.06153846153846,
      "grad_norm": 6.28125,
      "learning_rate": 2.208676496150427e-05,
      "loss": 0.0404,
      "step": 1499
    },
    {
      "epoch": 23.076923076923077,
      "grad_norm": 0.47265625,
      "learning_rate": 2.2081162585116872e-05,
      "loss": 0.0048,
      "step": 1500
    },
    {
      "epoch": 23.092307692307692,
      "grad_norm": 1.28125,
      "learning_rate": 2.2075555538987227e-05,
      "loss": 0.0131,
      "step": 1501
    },
    {
      "epoch": 23.107692307692307,
      "grad_norm": 4.625,
      "learning_rate": 2.2069943825848133e-05,
      "loss": 0.0284,
      "step": 1502
    },
    {
      "epoch": 23.123076923076923,
      "grad_norm": 2.5,
      "learning_rate": 2.206432744843468e-05,
      "loss": 0.0184,
      "step": 1503
    },
    {
      "epoch": 23.138461538461538,
      "grad_norm": 1.6953125,
      "learning_rate": 2.2058706409484213e-05,
      "loss": 0.0195,
      "step": 1504
    },
    {
      "epoch": 23.153846153846153,
      "grad_norm": 2.03125,
      "learning_rate": 2.2053080711736364e-05,
      "loss": 0.0143,
      "step": 1505
    },
    {
      "epoch": 23.16923076923077,
      "grad_norm": 1.421875,
      "learning_rate": 2.2047450357933032e-05,
      "loss": 0.0182,
      "step": 1506
    },
    {
      "epoch": 23.184615384615384,
      "grad_norm": 3.09375,
      "learning_rate": 2.2041815350818374e-05,
      "loss": 0.025,
      "step": 1507
    },
    {
      "epoch": 23.2,
      "grad_norm": 1.4296875,
      "learning_rate": 2.2036175693138827e-05,
      "loss": 0.0132,
      "step": 1508
    },
    {
      "epoch": 23.215384615384615,
      "grad_norm": 1.0,
      "learning_rate": 2.20305313876431e-05,
      "loss": 0.0101,
      "step": 1509
    },
    {
      "epoch": 23.23076923076923,
      "grad_norm": 4.15625,
      "learning_rate": 2.202488243708215e-05,
      "loss": 0.0398,
      "step": 1510
    },
    {
      "epoch": 23.246153846153845,
      "grad_norm": 5.375,
      "learning_rate": 2.201922884420921e-05,
      "loss": 0.0451,
      "step": 1511
    },
    {
      "epoch": 23.26153846153846,
      "grad_norm": 4.03125,
      "learning_rate": 2.2013570611779775e-05,
      "loss": 0.0782,
      "step": 1512
    },
    {
      "epoch": 23.276923076923076,
      "grad_norm": 0.80078125,
      "learning_rate": 2.200790774255159e-05,
      "loss": 0.0048,
      "step": 1513
    },
    {
      "epoch": 23.29230769230769,
      "grad_norm": 2.90625,
      "learning_rate": 2.2002240239284677e-05,
      "loss": 0.0249,
      "step": 1514
    },
    {
      "epoch": 23.307692307692307,
      "grad_norm": 2.609375,
      "learning_rate": 2.199656810474131e-05,
      "loss": 0.0201,
      "step": 1515
    },
    {
      "epoch": 23.323076923076922,
      "grad_norm": 1.65625,
      "learning_rate": 2.1990891341686008e-05,
      "loss": 0.0207,
      "step": 1516
    },
    {
      "epoch": 23.338461538461537,
      "grad_norm": 3.671875,
      "learning_rate": 2.1985209952885573e-05,
      "loss": 0.0197,
      "step": 1517
    },
    {
      "epoch": 23.353846153846153,
      "grad_norm": 0.58984375,
      "learning_rate": 2.1979523941109026e-05,
      "loss": 0.0254,
      "step": 1518
    },
    {
      "epoch": 23.369230769230768,
      "grad_norm": 3.828125,
      "learning_rate": 2.1973833309127666e-05,
      "loss": 0.0217,
      "step": 1519
    },
    {
      "epoch": 23.384615384615383,
      "grad_norm": 1.1953125,
      "learning_rate": 2.1968138059715048e-05,
      "loss": 0.009,
      "step": 1520
    },
    {
      "epoch": 23.4,
      "grad_norm": 4.53125,
      "learning_rate": 2.1962438195646958e-05,
      "loss": 0.0325,
      "step": 1521
    },
    {
      "epoch": 23.415384615384614,
      "grad_norm": 0.7109375,
      "learning_rate": 2.195673371970144e-05,
      "loss": 0.0238,
      "step": 1522
    },
    {
      "epoch": 23.43076923076923,
      "grad_norm": 1.0703125,
      "learning_rate": 2.195102463465879e-05,
      "loss": 0.0139,
      "step": 1523
    },
    {
      "epoch": 23.446153846153845,
      "grad_norm": 3.140625,
      "learning_rate": 2.1945310943301544e-05,
      "loss": 0.0194,
      "step": 1524
    },
    {
      "epoch": 23.46153846153846,
      "grad_norm": 4.8125,
      "learning_rate": 2.193959264841449e-05,
      "loss": 0.0771,
      "step": 1525
    },
    {
      "epoch": 23.476923076923075,
      "grad_norm": 1.15625,
      "learning_rate": 2.1933869752784654e-05,
      "loss": 0.0097,
      "step": 1526
    },
    {
      "epoch": 23.49230769230769,
      "grad_norm": 2.171875,
      "learning_rate": 2.1928142259201297e-05,
      "loss": 0.0593,
      "step": 1527
    },
    {
      "epoch": 23.50769230769231,
      "grad_norm": 4.625,
      "learning_rate": 2.1922410170455944e-05,
      "loss": 0.0326,
      "step": 1528
    },
    {
      "epoch": 23.523076923076925,
      "grad_norm": 2.46875,
      "learning_rate": 2.1916673489342337e-05,
      "loss": 0.028,
      "step": 1529
    },
    {
      "epoch": 23.53846153846154,
      "grad_norm": 5.625,
      "learning_rate": 2.1910932218656464e-05,
      "loss": 0.0457,
      "step": 1530
    },
    {
      "epoch": 23.553846153846155,
      "grad_norm": 2.078125,
      "learning_rate": 2.1905186361196556e-05,
      "loss": 0.0501,
      "step": 1531
    },
    {
      "epoch": 23.56923076923077,
      "grad_norm": 1.5859375,
      "learning_rate": 2.189943591976307e-05,
      "loss": 0.0107,
      "step": 1532
    },
    {
      "epoch": 23.584615384615386,
      "grad_norm": 0.7109375,
      "learning_rate": 2.1893680897158702e-05,
      "loss": 0.0441,
      "step": 1533
    },
    {
      "epoch": 23.6,
      "grad_norm": 1.703125,
      "learning_rate": 2.188792129618838e-05,
      "loss": 0.0475,
      "step": 1534
    },
    {
      "epoch": 23.615384615384617,
      "grad_norm": 0.4453125,
      "learning_rate": 2.1882157119659265e-05,
      "loss": 0.0206,
      "step": 1535
    },
    {
      "epoch": 23.630769230769232,
      "grad_norm": 1.46875,
      "learning_rate": 2.1876388370380745e-05,
      "loss": 0.0189,
      "step": 1536
    },
    {
      "epoch": 23.646153846153847,
      "grad_norm": 0.36328125,
      "learning_rate": 2.187061505116444e-05,
      "loss": 0.0369,
      "step": 1537
    },
    {
      "epoch": 23.661538461538463,
      "grad_norm": 2.75,
      "learning_rate": 2.186483716482419e-05,
      "loss": 0.0177,
      "step": 1538
    },
    {
      "epoch": 23.676923076923078,
      "grad_norm": 5.21875,
      "learning_rate": 2.1859054714176076e-05,
      "loss": 0.0554,
      "step": 1539
    },
    {
      "epoch": 23.692307692307693,
      "grad_norm": 4.15625,
      "learning_rate": 2.1853267702038392e-05,
      "loss": 0.0828,
      "step": 1540
    },
    {
      "epoch": 23.70769230769231,
      "grad_norm": 3.359375,
      "learning_rate": 2.184747613123165e-05,
      "loss": 0.0383,
      "step": 1541
    },
    {
      "epoch": 23.723076923076924,
      "grad_norm": 6.125,
      "learning_rate": 2.1841680004578596e-05,
      "loss": 0.0576,
      "step": 1542
    },
    {
      "epoch": 23.73846153846154,
      "grad_norm": 1.2109375,
      "learning_rate": 2.1835879324904195e-05,
      "loss": 0.011,
      "step": 1543
    },
    {
      "epoch": 23.753846153846155,
      "grad_norm": 1.375,
      "learning_rate": 2.183007409503562e-05,
      "loss": 0.0144,
      "step": 1544
    },
    {
      "epoch": 23.76923076923077,
      "grad_norm": 2.5,
      "learning_rate": 2.1824264317802278e-05,
      "loss": 0.0231,
      "step": 1545
    },
    {
      "epoch": 23.784615384615385,
      "grad_norm": 0.5546875,
      "learning_rate": 2.181844999603578e-05,
      "loss": 0.0223,
      "step": 1546
    },
    {
      "epoch": 23.8,
      "grad_norm": 2.59375,
      "learning_rate": 2.1812631132569953e-05,
      "loss": 0.0279,
      "step": 1547
    },
    {
      "epoch": 23.815384615384616,
      "grad_norm": 14.1875,
      "learning_rate": 2.1806807730240845e-05,
      "loss": 0.0354,
      "step": 1548
    },
    {
      "epoch": 23.83076923076923,
      "grad_norm": 1.046875,
      "learning_rate": 2.1800979791886707e-05,
      "loss": 0.0203,
      "step": 1549
    },
    {
      "epoch": 23.846153846153847,
      "grad_norm": 3.734375,
      "learning_rate": 2.1795147320348012e-05,
      "loss": 0.0731,
      "step": 1550
    },
    {
      "epoch": 23.861538461538462,
      "grad_norm": 3.796875,
      "learning_rate": 2.178931031846743e-05,
      "loss": 0.0283,
      "step": 1551
    },
    {
      "epoch": 23.876923076923077,
      "grad_norm": 4.125,
      "learning_rate": 2.1783468789089844e-05,
      "loss": 0.0236,
      "step": 1552
    },
    {
      "epoch": 23.892307692307693,
      "grad_norm": 1.78125,
      "learning_rate": 2.1777622735062354e-05,
      "loss": 0.0132,
      "step": 1553
    },
    {
      "epoch": 23.907692307692308,
      "grad_norm": 3.21875,
      "learning_rate": 2.1771772159234243e-05,
      "loss": 0.0416,
      "step": 1554
    },
    {
      "epoch": 23.923076923076923,
      "grad_norm": 2.78125,
      "learning_rate": 2.1765917064457013e-05,
      "loss": 0.0333,
      "step": 1555
    },
    {
      "epoch": 23.93846153846154,
      "grad_norm": 1.6171875,
      "learning_rate": 2.1760057453584376e-05,
      "loss": 0.0118,
      "step": 1556
    },
    {
      "epoch": 23.953846153846154,
      "grad_norm": 3.265625,
      "learning_rate": 2.175419332947223e-05,
      "loss": 0.0254,
      "step": 1557
    },
    {
      "epoch": 23.96923076923077,
      "grad_norm": 2.421875,
      "learning_rate": 2.174832469497867e-05,
      "loss": 0.0106,
      "step": 1558
    },
    {
      "epoch": 23.984615384615385,
      "grad_norm": 3.90625,
      "learning_rate": 2.1742451552964006e-05,
      "loss": 0.0089,
      "step": 1559
    },
    {
      "epoch": 24.0,
      "grad_norm": 0.96484375,
      "learning_rate": 2.1736573906290736e-05,
      "loss": 0.0097,
      "step": 1560
    },
    {
      "epoch": 24.015384615384615,
      "grad_norm": 0.65625,
      "learning_rate": 2.1730691757823553e-05,
      "loss": 0.018,
      "step": 1561
    },
    {
      "epoch": 24.03076923076923,
      "grad_norm": 3.0,
      "learning_rate": 2.1724805110429335e-05,
      "loss": 0.0143,
      "step": 1562
    },
    {
      "epoch": 24.046153846153846,
      "grad_norm": 0.86328125,
      "learning_rate": 2.1718913966977175e-05,
      "loss": 0.0475,
      "step": 1563
    },
    {
      "epoch": 24.06153846153846,
      "grad_norm": 2.109375,
      "learning_rate": 2.1713018330338343e-05,
      "loss": 0.0251,
      "step": 1564
    },
    {
      "epoch": 24.076923076923077,
      "grad_norm": 1.9375,
      "learning_rate": 2.1707118203386292e-05,
      "loss": 0.0354,
      "step": 1565
    },
    {
      "epoch": 24.092307692307692,
      "grad_norm": 4.625,
      "learning_rate": 2.1701213588996683e-05,
      "loss": 0.0444,
      "step": 1566
    },
    {
      "epoch": 24.107692307692307,
      "grad_norm": 1.71875,
      "learning_rate": 2.1695304490047344e-05,
      "loss": 0.0108,
      "step": 1567
    },
    {
      "epoch": 24.123076923076923,
      "grad_norm": 4.15625,
      "learning_rate": 2.1689390909418305e-05,
      "loss": 0.0267,
      "step": 1568
    },
    {
      "epoch": 24.138461538461538,
      "grad_norm": 0.40234375,
      "learning_rate": 2.168347284999177e-05,
      "loss": 0.0301,
      "step": 1569
    },
    {
      "epoch": 24.153846153846153,
      "grad_norm": 5.21875,
      "learning_rate": 2.1677550314652127e-05,
      "loss": 0.0334,
      "step": 1570
    },
    {
      "epoch": 24.16923076923077,
      "grad_norm": 3.828125,
      "learning_rate": 2.1671623306285956e-05,
      "loss": 0.0304,
      "step": 1571
    },
    {
      "epoch": 24.184615384615384,
      "grad_norm": 3.109375,
      "learning_rate": 2.1665691827782003e-05,
      "loss": 0.0285,
      "step": 1572
    },
    {
      "epoch": 24.2,
      "grad_norm": 0.49609375,
      "learning_rate": 2.1659755882031196e-05,
      "loss": 0.014,
      "step": 1573
    },
    {
      "epoch": 24.215384615384615,
      "grad_norm": 1.40625,
      "learning_rate": 2.1653815471926646e-05,
      "loss": 0.0095,
      "step": 1574
    },
    {
      "epoch": 24.23076923076923,
      "grad_norm": 0.8828125,
      "learning_rate": 2.164787060036364e-05,
      "loss": 0.0077,
      "step": 1575
    },
    {
      "epoch": 24.246153846153845,
      "grad_norm": 8.5,
      "learning_rate": 2.1641921270239632e-05,
      "loss": 0.0374,
      "step": 1576
    },
    {
      "epoch": 24.26153846153846,
      "grad_norm": 0.357421875,
      "learning_rate": 2.1635967484454257e-05,
      "loss": 0.0046,
      "step": 1577
    },
    {
      "epoch": 24.276923076923076,
      "grad_norm": 1.46875,
      "learning_rate": 2.163000924590931e-05,
      "loss": 0.0164,
      "step": 1578
    },
    {
      "epoch": 24.29230769230769,
      "grad_norm": 7.25,
      "learning_rate": 2.162404655750877e-05,
      "loss": 0.0933,
      "step": 1579
    },
    {
      "epoch": 24.307692307692307,
      "grad_norm": 1.2109375,
      "learning_rate": 2.161807942215878e-05,
      "loss": 0.0148,
      "step": 1580
    },
    {
      "epoch": 24.323076923076922,
      "grad_norm": 0.57421875,
      "learning_rate": 2.1612107842767647e-05,
      "loss": 0.0118,
      "step": 1581
    },
    {
      "epoch": 24.338461538461537,
      "grad_norm": 7.5,
      "learning_rate": 2.1606131822245844e-05,
      "loss": 0.0483,
      "step": 1582
    },
    {
      "epoch": 24.353846153846153,
      "grad_norm": 1.34375,
      "learning_rate": 2.1600151363506015e-05,
      "loss": 0.0112,
      "step": 1583
    },
    {
      "epoch": 24.369230769230768,
      "grad_norm": 3.421875,
      "learning_rate": 2.159416646946296e-05,
      "loss": 0.0475,
      "step": 1584
    },
    {
      "epoch": 24.384615384615383,
      "grad_norm": 3.109375,
      "learning_rate": 2.158817714303365e-05,
      "loss": 0.0252,
      "step": 1585
    },
    {
      "epoch": 24.4,
      "grad_norm": 4.03125,
      "learning_rate": 2.15821833871372e-05,
      "loss": 0.0267,
      "step": 1586
    },
    {
      "epoch": 24.415384615384614,
      "grad_norm": 7.4375,
      "learning_rate": 2.1576185204694898e-05,
      "loss": 0.0446,
      "step": 1587
    },
    {
      "epoch": 24.43076923076923,
      "grad_norm": 3.96875,
      "learning_rate": 2.157018259863019e-05,
      "loss": 0.0205,
      "step": 1588
    },
    {
      "epoch": 24.446153846153845,
      "grad_norm": 1.59375,
      "learning_rate": 2.1564175571868668e-05,
      "loss": 0.007,
      "step": 1589
    },
    {
      "epoch": 24.46153846153846,
      "grad_norm": 3.90625,
      "learning_rate": 2.1558164127338086e-05,
      "loss": 0.0118,
      "step": 1590
    },
    {
      "epoch": 24.476923076923075,
      "grad_norm": 0.80078125,
      "learning_rate": 2.1552148267968347e-05,
      "loss": 0.0369,
      "step": 1591
    },
    {
      "epoch": 24.49230769230769,
      "grad_norm": 8.375,
      "learning_rate": 2.1546127996691506e-05,
      "loss": 0.039,
      "step": 1592
    },
    {
      "epoch": 24.50769230769231,
      "grad_norm": 4.25,
      "learning_rate": 2.1540103316441777e-05,
      "loss": 0.0382,
      "step": 1593
    },
    {
      "epoch": 24.523076923076925,
      "grad_norm": 1.8828125,
      "learning_rate": 2.153407423015551e-05,
      "loss": 0.0135,
      "step": 1594
    },
    {
      "epoch": 24.53846153846154,
      "grad_norm": 3.171875,
      "learning_rate": 2.152804074077121e-05,
      "loss": 0.0199,
      "step": 1595
    },
    {
      "epoch": 24.553846153846155,
      "grad_norm": 2.4375,
      "learning_rate": 2.152200285122953e-05,
      "loss": 0.0203,
      "step": 1596
    },
    {
      "epoch": 24.56923076923077,
      "grad_norm": 3.921875,
      "learning_rate": 2.151596056447326e-05,
      "loss": 0.0181,
      "step": 1597
    },
    {
      "epoch": 24.584615384615386,
      "grad_norm": 2.984375,
      "learning_rate": 2.1509913883447335e-05,
      "loss": 0.0687,
      "step": 1598
    },
    {
      "epoch": 24.6,
      "grad_norm": 3.25,
      "learning_rate": 2.1503862811098837e-05,
      "loss": 0.0316,
      "step": 1599
    },
    {
      "epoch": 24.615384615384617,
      "grad_norm": 5.03125,
      "learning_rate": 2.1497807350376986e-05,
      "loss": 0.0209,
      "step": 1600
    },
    {
      "epoch": 24.630769230769232,
      "grad_norm": 3.296875,
      "learning_rate": 2.1491747504233138e-05,
      "loss": 0.0125,
      "step": 1601
    },
    {
      "epoch": 24.646153846153847,
      "grad_norm": 2.421875,
      "learning_rate": 2.1485683275620794e-05,
      "loss": 0.0338,
      "step": 1602
    },
    {
      "epoch": 24.661538461538463,
      "grad_norm": 2.796875,
      "learning_rate": 2.1479614667495578e-05,
      "loss": 0.0317,
      "step": 1603
    },
    {
      "epoch": 24.676923076923078,
      "grad_norm": 0.6796875,
      "learning_rate": 2.1473541682815257e-05,
      "loss": 0.015,
      "step": 1604
    },
    {
      "epoch": 24.692307692307693,
      "grad_norm": 1.5859375,
      "learning_rate": 2.1467464324539734e-05,
      "loss": 0.0084,
      "step": 1605
    },
    {
      "epoch": 24.70769230769231,
      "grad_norm": 6.09375,
      "learning_rate": 2.1461382595631036e-05,
      "loss": 0.0527,
      "step": 1606
    },
    {
      "epoch": 24.723076923076924,
      "grad_norm": 2.203125,
      "learning_rate": 2.1455296499053324e-05,
      "loss": 0.0266,
      "step": 1607
    },
    {
      "epoch": 24.73846153846154,
      "grad_norm": 6.375,
      "learning_rate": 2.144920603777289e-05,
      "loss": 0.1042,
      "step": 1608
    },
    {
      "epoch": 24.753846153846155,
      "grad_norm": 0.7265625,
      "learning_rate": 2.1443111214758148e-05,
      "loss": 0.0489,
      "step": 1609
    },
    {
      "epoch": 24.76923076923077,
      "grad_norm": 1.828125,
      "learning_rate": 2.1437012032979646e-05,
      "loss": 0.0093,
      "step": 1610
    },
    {
      "epoch": 24.784615384615385,
      "grad_norm": 1.2578125,
      "learning_rate": 2.1430908495410042e-05,
      "loss": 0.0096,
      "step": 1611
    },
    {
      "epoch": 24.8,
      "grad_norm": 1.1015625,
      "learning_rate": 2.1424800605024137e-05,
      "loss": 0.0193,
      "step": 1612
    },
    {
      "epoch": 24.815384615384616,
      "grad_norm": 3.078125,
      "learning_rate": 2.141868836479883e-05,
      "loss": 0.0388,
      "step": 1613
    },
    {
      "epoch": 24.83076923076923,
      "grad_norm": 0.72265625,
      "learning_rate": 2.1412571777713163e-05,
      "loss": 0.012,
      "step": 1614
    },
    {
      "epoch": 24.846153846153847,
      "grad_norm": 0.333984375,
      "learning_rate": 2.1406450846748282e-05,
      "loss": 0.0074,
      "step": 1615
    },
    {
      "epoch": 24.861538461538462,
      "grad_norm": 4.03125,
      "learning_rate": 2.140032557488746e-05,
      "loss": 0.0281,
      "step": 1616
    },
    {
      "epoch": 24.876923076923077,
      "grad_norm": 0.330078125,
      "learning_rate": 2.139419596511607e-05,
      "loss": 0.0126,
      "step": 1617
    },
    {
      "epoch": 24.892307692307693,
      "grad_norm": 2.515625,
      "learning_rate": 2.138806202042161e-05,
      "loss": 0.0274,
      "step": 1618
    },
    {
      "epoch": 24.907692307692308,
      "grad_norm": 2.6875,
      "learning_rate": 2.1381923743793704e-05,
      "loss": 0.0217,
      "step": 1619
    },
    {
      "epoch": 24.923076923076923,
      "grad_norm": 3.6875,
      "learning_rate": 2.1375781138224062e-05,
      "loss": 0.0364,
      "step": 1620
    },
    {
      "epoch": 24.93846153846154,
      "grad_norm": 0.8984375,
      "learning_rate": 2.136963420670651e-05,
      "loss": 0.0353,
      "step": 1621
    },
    {
      "epoch": 24.953846153846154,
      "grad_norm": 0.6953125,
      "learning_rate": 2.1363482952237e-05,
      "loss": 0.017,
      "step": 1622
    },
    {
      "epoch": 24.96923076923077,
      "grad_norm": 5.28125,
      "learning_rate": 2.135732737781358e-05,
      "loss": 0.0414,
      "step": 1623
    },
    {
      "epoch": 24.984615384615385,
      "grad_norm": 0.40625,
      "learning_rate": 2.1351167486436388e-05,
      "loss": 0.0046,
      "step": 1624
    },
    {
      "epoch": 25.0,
      "grad_norm": 2.375,
      "learning_rate": 2.1345003281107692e-05,
      "loss": 0.0245,
      "step": 1625
    },
    {
      "epoch": 25.015384615384615,
      "grad_norm": 0.60546875,
      "learning_rate": 2.1338834764831845e-05,
      "loss": 0.0178,
      "step": 1626
    },
    {
      "epoch": 25.03076923076923,
      "grad_norm": 5.46875,
      "learning_rate": 2.1332661940615312e-05,
      "loss": 0.0334,
      "step": 1627
    },
    {
      "epoch": 25.046153846153846,
      "grad_norm": 1.34375,
      "learning_rate": 2.1326484811466646e-05,
      "loss": 0.0173,
      "step": 1628
    },
    {
      "epoch": 25.06153846153846,
      "grad_norm": 0.76953125,
      "learning_rate": 2.132030338039651e-05,
      "loss": 0.0184,
      "step": 1629
    },
    {
      "epoch": 25.076923076923077,
      "grad_norm": 5.625,
      "learning_rate": 2.1314117650417657e-05,
      "loss": 0.044,
      "step": 1630
    },
    {
      "epoch": 25.092307692307692,
      "grad_norm": 1.0546875,
      "learning_rate": 2.1307927624544934e-05,
      "loss": 0.0504,
      "step": 1631
    },
    {
      "epoch": 25.107692307692307,
      "grad_norm": 2.515625,
      "learning_rate": 2.130173330579529e-05,
      "loss": 0.0113,
      "step": 1632
    },
    {
      "epoch": 25.123076923076923,
      "grad_norm": 5.15625,
      "learning_rate": 2.1295534697187756e-05,
      "loss": 0.0551,
      "step": 1633
    },
    {
      "epoch": 25.138461538461538,
      "grad_norm": 1.8515625,
      "learning_rate": 2.128933180174346e-05,
      "loss": 0.0386,
      "step": 1634
    },
    {
      "epoch": 25.153846153846153,
      "grad_norm": 1.5078125,
      "learning_rate": 2.1283124622485616e-05,
      "loss": 0.0061,
      "step": 1635
    },
    {
      "epoch": 25.16923076923077,
      "grad_norm": 2.921875,
      "learning_rate": 2.1276913162439532e-05,
      "loss": 0.0286,
      "step": 1636
    },
    {
      "epoch": 25.184615384615384,
      "grad_norm": 2.59375,
      "learning_rate": 2.1270697424632593e-05,
      "loss": 0.0331,
      "step": 1637
    },
    {
      "epoch": 25.2,
      "grad_norm": 3.0,
      "learning_rate": 2.126447741209428e-05,
      "loss": 0.0269,
      "step": 1638
    },
    {
      "epoch": 25.215384615384615,
      "grad_norm": 0.8203125,
      "learning_rate": 2.1258253127856148e-05,
      "loss": 0.0165,
      "step": 1639
    },
    {
      "epoch": 25.23076923076923,
      "grad_norm": 3.578125,
      "learning_rate": 2.1252024574951834e-05,
      "loss": 0.0538,
      "step": 1640
    },
    {
      "epoch": 25.246153846153845,
      "grad_norm": 2.59375,
      "learning_rate": 2.124579175641707e-05,
      "loss": 0.0263,
      "step": 1641
    },
    {
      "epoch": 25.26153846153846,
      "grad_norm": 2.71875,
      "learning_rate": 2.1239554675289645e-05,
      "loss": 0.0251,
      "step": 1642
    },
    {
      "epoch": 25.276923076923076,
      "grad_norm": 3.0625,
      "learning_rate": 2.123331333460944e-05,
      "loss": 0.0316,
      "step": 1643
    },
    {
      "epoch": 25.29230769230769,
      "grad_norm": 0.765625,
      "learning_rate": 2.122706773741841e-05,
      "loss": 0.0103,
      "step": 1644
    },
    {
      "epoch": 25.307692307692307,
      "grad_norm": 1.90625,
      "learning_rate": 2.122081788676058e-05,
      "loss": 0.0358,
      "step": 1645
    },
    {
      "epoch": 25.323076923076922,
      "grad_norm": 1.4765625,
      "learning_rate": 2.121456378568206e-05,
      "loss": 0.0213,
      "step": 1646
    },
    {
      "epoch": 25.338461538461537,
      "grad_norm": 1.453125,
      "learning_rate": 2.120830543723101e-05,
      "loss": 0.0275,
      "step": 1647
    },
    {
      "epoch": 25.353846153846153,
      "grad_norm": 0.72265625,
      "learning_rate": 2.1202042844457682e-05,
      "loss": 0.0306,
      "step": 1648
    },
    {
      "epoch": 25.369230769230768,
      "grad_norm": 0.78515625,
      "learning_rate": 2.1195776010414388e-05,
      "loss": 0.025,
      "step": 1649
    },
    {
      "epoch": 25.384615384615383,
      "grad_norm": 0.3359375,
      "learning_rate": 2.11895049381555e-05,
      "loss": 0.0142,
      "step": 1650
    },
    {
      "epoch": 25.4,
      "grad_norm": 4.03125,
      "learning_rate": 2.1183229630737467e-05,
      "loss": 0.0762,
      "step": 1651
    },
    {
      "epoch": 25.415384615384614,
      "grad_norm": 2.484375,
      "learning_rate": 2.11769500912188e-05,
      "loss": 0.0257,
      "step": 1652
    },
    {
      "epoch": 25.43076923076923,
      "grad_norm": 1.1640625,
      "learning_rate": 2.117066632266006e-05,
      "loss": 0.0242,
      "step": 1653
    },
    {
      "epoch": 25.446153846153845,
      "grad_norm": 2.53125,
      "learning_rate": 2.1164378328123898e-05,
      "loss": 0.0102,
      "step": 1654
    },
    {
      "epoch": 25.46153846153846,
      "grad_norm": 3.765625,
      "learning_rate": 2.1158086110674987e-05,
      "loss": 0.0516,
      "step": 1655
    },
    {
      "epoch": 25.476923076923075,
      "grad_norm": 1.21875,
      "learning_rate": 2.1151789673380086e-05,
      "loss": 0.0224,
      "step": 1656
    },
    {
      "epoch": 25.49230769230769,
      "grad_norm": 2.9375,
      "learning_rate": 2.1145489019308e-05,
      "loss": 0.024,
      "step": 1657
    },
    {
      "epoch": 25.50769230769231,
      "grad_norm": 2.09375,
      "learning_rate": 2.11391841515296e-05,
      "loss": 0.0138,
      "step": 1658
    },
    {
      "epoch": 25.523076923076925,
      "grad_norm": 2.015625,
      "learning_rate": 2.1132875073117785e-05,
      "loss": 0.0374,
      "step": 1659
    },
    {
      "epoch": 25.53846153846154,
      "grad_norm": 3.765625,
      "learning_rate": 2.112656178714754e-05,
      "loss": 0.0326,
      "step": 1660
    },
    {
      "epoch": 25.553846153846155,
      "grad_norm": 4.5625,
      "learning_rate": 2.1120244296695874e-05,
      "loss": 0.0398,
      "step": 1661
    },
    {
      "epoch": 25.56923076923077,
      "grad_norm": 2.375,
      "learning_rate": 2.1113922604841858e-05,
      "loss": 0.0195,
      "step": 1662
    },
    {
      "epoch": 25.584615384615386,
      "grad_norm": 0.9140625,
      "learning_rate": 2.1107596714666604e-05,
      "loss": 0.0326,
      "step": 1663
    },
    {
      "epoch": 25.6,
      "grad_norm": 3.515625,
      "learning_rate": 2.1101266629253286e-05,
      "loss": 0.0464,
      "step": 1664
    },
    {
      "epoch": 25.615384615384617,
      "grad_norm": 2.09375,
      "learning_rate": 2.1094932351687095e-05,
      "loss": 0.0176,
      "step": 1665
    },
    {
      "epoch": 25.630769230769232,
      "grad_norm": 2.5,
      "learning_rate": 2.1088593885055288e-05,
      "loss": 0.0265,
      "step": 1666
    },
    {
      "epoch": 25.646153846153847,
      "grad_norm": 5.59375,
      "learning_rate": 2.1082251232447158e-05,
      "loss": 0.051,
      "step": 1667
    },
    {
      "epoch": 25.661538461538463,
      "grad_norm": 0.439453125,
      "learning_rate": 2.107590439695404e-05,
      "loss": 0.0134,
      "step": 1668
    },
    {
      "epoch": 25.676923076923078,
      "grad_norm": 5.5625,
      "learning_rate": 2.1069553381669296e-05,
      "loss": 0.0451,
      "step": 1669
    },
    {
      "epoch": 25.692307692307693,
      "grad_norm": 3.640625,
      "learning_rate": 2.106319818968834e-05,
      "loss": 0.042,
      "step": 1670
    },
    {
      "epoch": 25.70769230769231,
      "grad_norm": 1.5234375,
      "learning_rate": 2.105683882410861e-05,
      "loss": 0.0147,
      "step": 1671
    },
    {
      "epoch": 25.723076923076924,
      "grad_norm": 2.375,
      "learning_rate": 2.1050475288029586e-05,
      "loss": 0.0168,
      "step": 1672
    },
    {
      "epoch": 25.73846153846154,
      "grad_norm": 0.2294921875,
      "learning_rate": 2.1044107584552785e-05,
      "loss": 0.0421,
      "step": 1673
    },
    {
      "epoch": 25.753846153846155,
      "grad_norm": 4.65625,
      "learning_rate": 2.1037735716781736e-05,
      "loss": 0.0306,
      "step": 1674
    },
    {
      "epoch": 25.76923076923077,
      "grad_norm": 2.859375,
      "learning_rate": 2.1031359687822023e-05,
      "loss": 0.0504,
      "step": 1675
    },
    {
      "epoch": 25.784615384615385,
      "grad_norm": 1.1953125,
      "learning_rate": 2.1024979500781232e-05,
      "loss": 0.0149,
      "step": 1676
    },
    {
      "epoch": 25.8,
      "grad_norm": 0.734375,
      "learning_rate": 2.1018595158769e-05,
      "loss": 0.0385,
      "step": 1677
    },
    {
      "epoch": 25.815384615384616,
      "grad_norm": 1.0546875,
      "learning_rate": 2.101220666489697e-05,
      "loss": 0.0154,
      "step": 1678
    },
    {
      "epoch": 25.83076923076923,
      "grad_norm": 4.25,
      "learning_rate": 2.1005814022278816e-05,
      "loss": 0.025,
      "step": 1679
    },
    {
      "epoch": 25.846153846153847,
      "grad_norm": 6.78125,
      "learning_rate": 2.099941723403024e-05,
      "loss": 0.0473,
      "step": 1680
    },
    {
      "epoch": 25.861538461538462,
      "grad_norm": 3.765625,
      "learning_rate": 2.099301630326896e-05,
      "loss": 0.0225,
      "step": 1681
    },
    {
      "epoch": 25.876923076923077,
      "grad_norm": 1.390625,
      "learning_rate": 2.0986611233114702e-05,
      "loss": 0.0168,
      "step": 1682
    },
    {
      "epoch": 25.892307692307693,
      "grad_norm": 0.29296875,
      "learning_rate": 2.0980202026689227e-05,
      "loss": 0.0238,
      "step": 1683
    },
    {
      "epoch": 25.907692307692308,
      "grad_norm": 3.59375,
      "learning_rate": 2.0973788687116304e-05,
      "loss": 0.0224,
      "step": 1684
    },
    {
      "epoch": 25.923076923076923,
      "grad_norm": 2.5,
      "learning_rate": 2.0967371217521716e-05,
      "loss": 0.0147,
      "step": 1685
    },
    {
      "epoch": 25.93846153846154,
      "grad_norm": 1.984375,
      "learning_rate": 2.096094962103326e-05,
      "loss": 0.0344,
      "step": 1686
    },
    {
      "epoch": 25.953846153846154,
      "grad_norm": 1.53125,
      "learning_rate": 2.0954523900780742e-05,
      "loss": 0.0205,
      "step": 1687
    },
    {
      "epoch": 25.96923076923077,
      "grad_norm": 3.4375,
      "learning_rate": 2.0948094059895984e-05,
      "loss": 0.0212,
      "step": 1688
    },
    {
      "epoch": 25.984615384615385,
      "grad_norm": 0.81640625,
      "learning_rate": 2.0941660101512806e-05,
      "loss": 0.0155,
      "step": 1689
    },
    {
      "epoch": 26.0,
      "grad_norm": 3.75,
      "learning_rate": 2.0935222028767047e-05,
      "loss": 0.056,
      "step": 1690
    },
    {
      "epoch": 26.015384615384615,
      "grad_norm": 2.109375,
      "learning_rate": 2.092877984479654e-05,
      "loss": 0.0141,
      "step": 1691
    },
    {
      "epoch": 26.03076923076923,
      "grad_norm": 0.71484375,
      "learning_rate": 2.0922333552741136e-05,
      "loss": 0.0061,
      "step": 1692
    },
    {
      "epoch": 26.046153846153846,
      "grad_norm": 0.57421875,
      "learning_rate": 2.091588315574267e-05,
      "loss": 0.036,
      "step": 1693
    },
    {
      "epoch": 26.06153846153846,
      "grad_norm": 5.25,
      "learning_rate": 2.0909428656944995e-05,
      "loss": 0.0378,
      "step": 1694
    },
    {
      "epoch": 26.076923076923077,
      "grad_norm": 0.515625,
      "learning_rate": 2.0902970059493955e-05,
      "loss": 0.0056,
      "step": 1695
    },
    {
      "epoch": 26.092307692307692,
      "grad_norm": 4.1875,
      "learning_rate": 2.089650736653738e-05,
      "loss": 0.0229,
      "step": 1696
    },
    {
      "epoch": 26.107692307692307,
      "grad_norm": 5.3125,
      "learning_rate": 2.089004058122512e-05,
      "loss": 0.0391,
      "step": 1697
    },
    {
      "epoch": 26.123076923076923,
      "grad_norm": 0.87890625,
      "learning_rate": 2.0883569706709e-05,
      "loss": 0.0229,
      "step": 1698
    },
    {
      "epoch": 26.138461538461538,
      "grad_norm": 2.21875,
      "learning_rate": 2.0877094746142857e-05,
      "loss": 0.0336,
      "step": 1699
    },
    {
      "epoch": 26.153846153846153,
      "grad_norm": 10.6875,
      "learning_rate": 2.0870615702682494e-05,
      "loss": 0.0439,
      "step": 1700
    },
    {
      "epoch": 26.16923076923077,
      "grad_norm": 1.4921875,
      "learning_rate": 2.086413257948573e-05,
      "loss": 0.0139,
      "step": 1701
    },
    {
      "epoch": 26.184615384615384,
      "grad_norm": 0.8359375,
      "learning_rate": 2.0857645379712353e-05,
      "loss": 0.0059,
      "step": 1702
    },
    {
      "epoch": 26.2,
      "grad_norm": 5.96875,
      "learning_rate": 2.0851154106524146e-05,
      "loss": 0.0226,
      "step": 1703
    },
    {
      "epoch": 26.215384615384615,
      "grad_norm": 0.66015625,
      "learning_rate": 2.0844658763084883e-05,
      "loss": 0.0129,
      "step": 1704
    },
    {
      "epoch": 26.23076923076923,
      "grad_norm": 0.1640625,
      "learning_rate": 2.0838159352560308e-05,
      "loss": 0.0036,
      "step": 1705
    },
    {
      "epoch": 26.246153846153845,
      "grad_norm": 2.296875,
      "learning_rate": 2.0831655878118155e-05,
      "loss": 0.0177,
      "step": 1706
    },
    {
      "epoch": 26.26153846153846,
      "grad_norm": 3.21875,
      "learning_rate": 2.0825148342928146e-05,
      "loss": 0.0236,
      "step": 1707
    },
    {
      "epoch": 26.276923076923076,
      "grad_norm": 1.2890625,
      "learning_rate": 2.081863675016197e-05,
      "loss": 0.0065,
      "step": 1708
    },
    {
      "epoch": 26.29230769230769,
      "grad_norm": 4.0625,
      "learning_rate": 2.081212110299329e-05,
      "loss": 0.0312,
      "step": 1709
    },
    {
      "epoch": 26.307692307692307,
      "grad_norm": 2.828125,
      "learning_rate": 2.080560140459777e-05,
      "loss": 0.0251,
      "step": 1710
    },
    {
      "epoch": 26.323076923076922,
      "grad_norm": 0.90625,
      "learning_rate": 2.0799077658153022e-05,
      "loss": 0.0078,
      "step": 1711
    },
    {
      "epoch": 26.338461538461537,
      "grad_norm": 2.8125,
      "learning_rate": 2.079254986683864e-05,
      "loss": 0.025,
      "step": 1712
    },
    {
      "epoch": 26.353846153846153,
      "grad_norm": 2.0,
      "learning_rate": 2.078601803383619e-05,
      "loss": 0.0155,
      "step": 1713
    },
    {
      "epoch": 26.369230769230768,
      "grad_norm": 6.21875,
      "learning_rate": 2.077948216232921e-05,
      "loss": 0.0449,
      "step": 1714
    },
    {
      "epoch": 26.384615384615383,
      "grad_norm": 1.625,
      "learning_rate": 2.077294225550321e-05,
      "loss": 0.0765,
      "step": 1715
    },
    {
      "epoch": 26.4,
      "grad_norm": 2.8125,
      "learning_rate": 2.0766398316545648e-05,
      "loss": 0.034,
      "step": 1716
    },
    {
      "epoch": 26.415384615384614,
      "grad_norm": 1.6328125,
      "learning_rate": 2.0759850348645972e-05,
      "loss": 0.0564,
      "step": 1717
    },
    {
      "epoch": 26.43076923076923,
      "grad_norm": 1.3046875,
      "learning_rate": 2.075329835499558e-05,
      "loss": 0.0109,
      "step": 1718
    },
    {
      "epoch": 26.446153846153845,
      "grad_norm": 0.271484375,
      "learning_rate": 2.074674233878783e-05,
      "loss": 0.0069,
      "step": 1719
    },
    {
      "epoch": 26.46153846153846,
      "grad_norm": 3.765625,
      "learning_rate": 2.074018230321804e-05,
      "loss": 0.0163,
      "step": 1720
    },
    {
      "epoch": 26.476923076923075,
      "grad_norm": 6.6875,
      "learning_rate": 2.0733618251483506e-05,
      "loss": 0.0358,
      "step": 1721
    },
    {
      "epoch": 26.49230769230769,
      "grad_norm": 1.8203125,
      "learning_rate": 2.072705018678346e-05,
      "loss": 0.0337,
      "step": 1722
    },
    {
      "epoch": 26.50769230769231,
      "grad_norm": 1.96875,
      "learning_rate": 2.0720478112319092e-05,
      "loss": 0.0502,
      "step": 1723
    },
    {
      "epoch": 26.523076923076925,
      "grad_norm": 2.359375,
      "learning_rate": 2.0713902031293558e-05,
      "loss": 0.0673,
      "step": 1724
    },
    {
      "epoch": 26.53846153846154,
      "grad_norm": 0.2294921875,
      "learning_rate": 2.0707321946911957e-05,
      "loss": 0.0082,
      "step": 1725
    },
    {
      "epoch": 26.553846153846155,
      "grad_norm": 0.341796875,
      "learning_rate": 2.070073786238134e-05,
      "loss": 0.0053,
      "step": 1726
    },
    {
      "epoch": 26.56923076923077,
      "grad_norm": 1.875,
      "learning_rate": 2.069414978091072e-05,
      "loss": 0.0139,
      "step": 1727
    },
    {
      "epoch": 26.584615384615386,
      "grad_norm": 2.4375,
      "learning_rate": 2.0687557705711033e-05,
      "loss": 0.0224,
      "step": 1728
    },
    {
      "epoch": 26.6,
      "grad_norm": 1.25,
      "learning_rate": 2.0680961639995185e-05,
      "loss": 0.0262,
      "step": 1729
    },
    {
      "epoch": 26.615384615384617,
      "grad_norm": 4.03125,
      "learning_rate": 2.067436158697802e-05,
      "loss": 0.0739,
      "step": 1730
    },
    {
      "epoch": 26.630769230769232,
      "grad_norm": 7.21875,
      "learning_rate": 2.066775754987632e-05,
      "loss": 0.0486,
      "step": 1731
    },
    {
      "epoch": 26.646153846153847,
      "grad_norm": 5.71875,
      "learning_rate": 2.0661149531908812e-05,
      "loss": 0.0461,
      "step": 1732
    },
    {
      "epoch": 26.661538461538463,
      "grad_norm": 5.0625,
      "learning_rate": 2.0654537536296164e-05,
      "loss": 0.0386,
      "step": 1733
    },
    {
      "epoch": 26.676923076923078,
      "grad_norm": 2.5625,
      "learning_rate": 2.0647921566260987e-05,
      "loss": 0.0378,
      "step": 1734
    },
    {
      "epoch": 26.692307692307693,
      "grad_norm": 1.0625,
      "learning_rate": 2.064130162502782e-05,
      "loss": 0.0477,
      "step": 1735
    },
    {
      "epoch": 26.70769230769231,
      "grad_norm": 1.4296875,
      "learning_rate": 2.0634677715823137e-05,
      "loss": 0.0159,
      "step": 1736
    },
    {
      "epoch": 26.723076923076924,
      "grad_norm": 1.03125,
      "learning_rate": 2.062804984187536e-05,
      "loss": 0.0199,
      "step": 1737
    },
    {
      "epoch": 26.73846153846154,
      "grad_norm": 3.5,
      "learning_rate": 2.0621418006414833e-05,
      "loss": 0.0218,
      "step": 1738
    },
    {
      "epoch": 26.753846153846155,
      "grad_norm": 2.15625,
      "learning_rate": 2.0614782212673827e-05,
      "loss": 0.0277,
      "step": 1739
    },
    {
      "epoch": 26.76923076923077,
      "grad_norm": 3.015625,
      "learning_rate": 2.0608142463886553e-05,
      "loss": 0.043,
      "step": 1740
    },
    {
      "epoch": 26.784615384615385,
      "grad_norm": 2.125,
      "learning_rate": 2.0601498763289138e-05,
      "loss": 0.0519,
      "step": 1741
    },
    {
      "epoch": 26.8,
      "grad_norm": 1.8515625,
      "learning_rate": 2.0594851114119645e-05,
      "loss": 0.0276,
      "step": 1742
    },
    {
      "epoch": 26.815384615384616,
      "grad_norm": 0.6328125,
      "learning_rate": 2.0588199519618058e-05,
      "loss": 0.0069,
      "step": 1743
    },
    {
      "epoch": 26.83076923076923,
      "grad_norm": 2.8125,
      "learning_rate": 2.0581543983026276e-05,
      "loss": 0.058,
      "step": 1744
    },
    {
      "epoch": 26.846153846153847,
      "grad_norm": 1.203125,
      "learning_rate": 2.0574884507588137e-05,
      "loss": 0.0328,
      "step": 1745
    },
    {
      "epoch": 26.861538461538462,
      "grad_norm": 3.640625,
      "learning_rate": 2.0568221096549384e-05,
      "loss": 0.0335,
      "step": 1746
    },
    {
      "epoch": 26.876923076923077,
      "grad_norm": 3.96875,
      "learning_rate": 2.056155375315768e-05,
      "loss": 0.0276,
      "step": 1747
    },
    {
      "epoch": 26.892307692307693,
      "grad_norm": 2.125,
      "learning_rate": 2.055488248066261e-05,
      "loss": 0.0271,
      "step": 1748
    },
    {
      "epoch": 26.907692307692308,
      "grad_norm": 2.546875,
      "learning_rate": 2.0548207282315675e-05,
      "loss": 0.0482,
      "step": 1749
    },
    {
      "epoch": 26.923076923076923,
      "grad_norm": 2.140625,
      "learning_rate": 2.0541528161370276e-05,
      "loss": 0.0162,
      "step": 1750
    },
    {
      "epoch": 26.93846153846154,
      "grad_norm": 4.6875,
      "learning_rate": 2.0534845121081742e-05,
      "loss": 0.041,
      "step": 1751
    },
    {
      "epoch": 26.953846153846154,
      "grad_norm": 0.50390625,
      "learning_rate": 2.0528158164707308e-05,
      "loss": 0.0224,
      "step": 1752
    },
    {
      "epoch": 26.96923076923077,
      "grad_norm": 5.84375,
      "learning_rate": 2.052146729550611e-05,
      "loss": 0.0363,
      "step": 1753
    },
    {
      "epoch": 26.984615384615385,
      "grad_norm": 1.5859375,
      "learning_rate": 2.0514772516739194e-05,
      "loss": 0.0115,
      "step": 1754
    },
    {
      "epoch": 27.0,
      "grad_norm": 4.0625,
      "learning_rate": 2.050807383166952e-05,
      "loss": 0.0569,
      "step": 1755
    },
    {
      "epoch": 27.015384615384615,
      "grad_norm": 3.4375,
      "learning_rate": 2.0501371243561946e-05,
      "loss": 0.0419,
      "step": 1756
    },
    {
      "epoch": 27.03076923076923,
      "grad_norm": 1.671875,
      "learning_rate": 2.0494664755683226e-05,
      "loss": 0.0192,
      "step": 1757
    },
    {
      "epoch": 27.046153846153846,
      "grad_norm": 5.78125,
      "learning_rate": 2.0487954371302027e-05,
      "loss": 0.0559,
      "step": 1758
    },
    {
      "epoch": 27.06153846153846,
      "grad_norm": 3.9375,
      "learning_rate": 2.0481240093688904e-05,
      "loss": 0.0678,
      "step": 1759
    },
    {
      "epoch": 27.076923076923077,
      "grad_norm": 1.6484375,
      "learning_rate": 2.047452192611632e-05,
      "loss": 0.0191,
      "step": 1760
    },
    {
      "epoch": 27.092307692307692,
      "grad_norm": 2.078125,
      "learning_rate": 2.0467799871858624e-05,
      "loss": 0.0587,
      "step": 1761
    },
    {
      "epoch": 27.107692307692307,
      "grad_norm": 1.6796875,
      "learning_rate": 2.0461073934192064e-05,
      "loss": 0.0597,
      "step": 1762
    },
    {
      "epoch": 27.123076923076923,
      "grad_norm": 1.2421875,
      "learning_rate": 2.045434411639478e-05,
      "loss": 0.0123,
      "step": 1763
    },
    {
      "epoch": 27.138461538461538,
      "grad_norm": 3.609375,
      "learning_rate": 2.0447610421746803e-05,
      "loss": 0.0308,
      "step": 1764
    },
    {
      "epoch": 27.153846153846153,
      "grad_norm": 0.6796875,
      "learning_rate": 2.0440872853530055e-05,
      "loss": 0.0156,
      "step": 1765
    },
    {
      "epoch": 27.16923076923077,
      "grad_norm": 2.09375,
      "learning_rate": 2.0434131415028346e-05,
      "loss": 0.0137,
      "step": 1766
    },
    {
      "epoch": 27.184615384615384,
      "grad_norm": 2.953125,
      "learning_rate": 2.042738610952737e-05,
      "loss": 0.0376,
      "step": 1767
    },
    {
      "epoch": 27.2,
      "grad_norm": 2.390625,
      "learning_rate": 2.0420636940314708e-05,
      "loss": 0.0335,
      "step": 1768
    },
    {
      "epoch": 27.215384615384615,
      "grad_norm": 0.482421875,
      "learning_rate": 2.041388391067982e-05,
      "loss": 0.0068,
      "step": 1769
    },
    {
      "epoch": 27.23076923076923,
      "grad_norm": 3.265625,
      "learning_rate": 2.0407127023914058e-05,
      "loss": 0.0614,
      "step": 1770
    },
    {
      "epoch": 27.246153846153845,
      "grad_norm": 1.140625,
      "learning_rate": 2.0400366283310636e-05,
      "loss": 0.0112,
      "step": 1771
    },
    {
      "epoch": 27.26153846153846,
      "grad_norm": 0.96875,
      "learning_rate": 2.0393601692164665e-05,
      "loss": 0.0238,
      "step": 1772
    },
    {
      "epoch": 27.276923076923076,
      "grad_norm": 3.96875,
      "learning_rate": 2.038683325377312e-05,
      "loss": 0.0252,
      "step": 1773
    },
    {
      "epoch": 27.29230769230769,
      "grad_norm": 4.75,
      "learning_rate": 2.0380060971434862e-05,
      "loss": 0.03,
      "step": 1774
    },
    {
      "epoch": 27.307692307692307,
      "grad_norm": 3.6875,
      "learning_rate": 2.037328484845061e-05,
      "loss": 0.0446,
      "step": 1775
    },
    {
      "epoch": 27.323076923076922,
      "grad_norm": 2.65625,
      "learning_rate": 2.036650488812297e-05,
      "loss": 0.0478,
      "step": 1776
    },
    {
      "epoch": 27.338461538461537,
      "grad_norm": 0.81640625,
      "learning_rate": 2.035972109375641e-05,
      "loss": 0.0248,
      "step": 1777
    },
    {
      "epoch": 27.353846153846153,
      "grad_norm": 2.140625,
      "learning_rate": 2.035293346865727e-05,
      "loss": 0.0283,
      "step": 1778
    },
    {
      "epoch": 27.369230769230768,
      "grad_norm": 1.5390625,
      "learning_rate": 2.0346142016133758e-05,
      "loss": 0.0349,
      "step": 1779
    },
    {
      "epoch": 27.384615384615383,
      "grad_norm": 3.53125,
      "learning_rate": 2.033934673949594e-05,
      "loss": 0.0394,
      "step": 1780
    },
    {
      "epoch": 27.4,
      "grad_norm": 1.2890625,
      "learning_rate": 2.033254764205576e-05,
      "loss": 0.0205,
      "step": 1781
    },
    {
      "epoch": 27.415384615384614,
      "grad_norm": 1.25,
      "learning_rate": 2.0325744727127005e-05,
      "loss": 0.0161,
      "step": 1782
    },
    {
      "epoch": 27.43076923076923,
      "grad_norm": 4.75,
      "learning_rate": 2.031893799802534e-05,
      "loss": 0.0214,
      "step": 1783
    },
    {
      "epoch": 27.446153846153845,
      "grad_norm": 1.59375,
      "learning_rate": 2.031212745806828e-05,
      "loss": 0.0146,
      "step": 1784
    },
    {
      "epoch": 27.46153846153846,
      "grad_norm": 1.5390625,
      "learning_rate": 2.0305313110575197e-05,
      "loss": 0.0088,
      "step": 1785
    },
    {
      "epoch": 27.476923076923075,
      "grad_norm": 2.515625,
      "learning_rate": 2.029849495886733e-05,
      "loss": 0.0319,
      "step": 1786
    },
    {
      "epoch": 27.49230769230769,
      "grad_norm": 0.890625,
      "learning_rate": 2.0291673006267753e-05,
      "loss": 0.018,
      "step": 1787
    },
    {
      "epoch": 27.50769230769231,
      "grad_norm": 2.453125,
      "learning_rate": 2.028484725610141e-05,
      "loss": 0.0389,
      "step": 1788
    },
    {
      "epoch": 27.523076923076925,
      "grad_norm": 2.28125,
      "learning_rate": 2.027801771169508e-05,
      "loss": 0.0154,
      "step": 1789
    },
    {
      "epoch": 27.53846153846154,
      "grad_norm": 0.1640625,
      "learning_rate": 2.0271184376377413e-05,
      "loss": 0.0153,
      "step": 1790
    },
    {
      "epoch": 27.553846153846155,
      "grad_norm": 3.515625,
      "learning_rate": 2.026434725347888e-05,
      "loss": 0.0121,
      "step": 1791
    },
    {
      "epoch": 27.56923076923077,
      "grad_norm": 1.0859375,
      "learning_rate": 2.0257506346331823e-05,
      "loss": 0.0216,
      "step": 1792
    },
    {
      "epoch": 27.584615384615386,
      "grad_norm": 2.3125,
      "learning_rate": 2.025066165827041e-05,
      "loss": 0.0262,
      "step": 1793
    },
    {
      "epoch": 27.6,
      "grad_norm": 10.1875,
      "learning_rate": 2.024381319263066e-05,
      "loss": 0.0869,
      "step": 1794
    },
    {
      "epoch": 27.615384615384617,
      "grad_norm": 0.8359375,
      "learning_rate": 2.0236960952750435e-05,
      "loss": 0.0059,
      "step": 1795
    },
    {
      "epoch": 27.630769230769232,
      "grad_norm": 4.0,
      "learning_rate": 2.0230104941969426e-05,
      "loss": 0.0494,
      "step": 1796
    },
    {
      "epoch": 27.646153846153847,
      "grad_norm": 4.3125,
      "learning_rate": 2.022324516362918e-05,
      "loss": 0.1094,
      "step": 1797
    },
    {
      "epoch": 27.661538461538463,
      "grad_norm": 1.484375,
      "learning_rate": 2.0216381621073064e-05,
      "loss": 0.0356,
      "step": 1798
    },
    {
      "epoch": 27.676923076923078,
      "grad_norm": 0.279296875,
      "learning_rate": 2.020951431764628e-05,
      "loss": 0.0148,
      "step": 1799
    },
    {
      "epoch": 27.692307692307693,
      "grad_norm": 0.2490234375,
      "learning_rate": 2.020264325669588e-05,
      "loss": 0.0124,
      "step": 1800
    },
    {
      "epoch": 27.70769230769231,
      "grad_norm": 2.328125,
      "learning_rate": 2.0195768441570727e-05,
      "loss": 0.0244,
      "step": 1801
    },
    {
      "epoch": 27.723076923076924,
      "grad_norm": 2.75,
      "learning_rate": 2.018888987562153e-05,
      "loss": 0.0223,
      "step": 1802
    },
    {
      "epoch": 27.73846153846154,
      "grad_norm": 4.0625,
      "learning_rate": 2.0182007562200805e-05,
      "loss": 0.0186,
      "step": 1803
    },
    {
      "epoch": 27.753846153846155,
      "grad_norm": 7.53125,
      "learning_rate": 2.017512150466292e-05,
      "loss": 0.0491,
      "step": 1804
    },
    {
      "epoch": 27.76923076923077,
      "grad_norm": 2.71875,
      "learning_rate": 2.0168231706364054e-05,
      "loss": 0.0293,
      "step": 1805
    },
    {
      "epoch": 27.784615384615385,
      "grad_norm": 3.921875,
      "learning_rate": 2.0161338170662208e-05,
      "loss": 0.064,
      "step": 1806
    },
    {
      "epoch": 27.8,
      "grad_norm": 2.1875,
      "learning_rate": 2.0154440900917206e-05,
      "loss": 0.033,
      "step": 1807
    },
    {
      "epoch": 27.815384615384616,
      "grad_norm": 3.84375,
      "learning_rate": 2.0147539900490697e-05,
      "loss": 0.0217,
      "step": 1808
    },
    {
      "epoch": 27.83076923076923,
      "grad_norm": 1.9375,
      "learning_rate": 2.0140635172746146e-05,
      "loss": 0.0219,
      "step": 1809
    },
    {
      "epoch": 27.846153846153847,
      "grad_norm": 0.8359375,
      "learning_rate": 2.0133726721048832e-05,
      "loss": 0.0227,
      "step": 1810
    },
    {
      "epoch": 27.861538461538462,
      "grad_norm": 0.5703125,
      "learning_rate": 2.012681454876585e-05,
      "loss": 0.025,
      "step": 1811
    },
    {
      "epoch": 27.876923076923077,
      "grad_norm": 1.5546875,
      "learning_rate": 2.01198986592661e-05,
      "loss": 0.0121,
      "step": 1812
    },
    {
      "epoch": 27.892307692307693,
      "grad_norm": 1.671875,
      "learning_rate": 2.011297905592032e-05,
      "loss": 0.031,
      "step": 1813
    },
    {
      "epoch": 27.907692307692308,
      "grad_norm": 2.484375,
      "learning_rate": 2.0106055742101026e-05,
      "loss": 0.0253,
      "step": 1814
    },
    {
      "epoch": 27.923076923076923,
      "grad_norm": 1.2421875,
      "learning_rate": 2.009912872118257e-05,
      "loss": 0.034,
      "step": 1815
    },
    {
      "epoch": 27.93846153846154,
      "grad_norm": 7.4375,
      "learning_rate": 2.0092197996541085e-05,
      "loss": 0.043,
      "step": 1816
    },
    {
      "epoch": 27.953846153846154,
      "grad_norm": 3.46875,
      "learning_rate": 2.008526357155453e-05,
      "loss": 0.0281,
      "step": 1817
    },
    {
      "epoch": 27.96923076923077,
      "grad_norm": 3.4375,
      "learning_rate": 2.0078325449602655e-05,
      "loss": 0.0286,
      "step": 1818
    },
    {
      "epoch": 27.984615384615385,
      "grad_norm": 1.7890625,
      "learning_rate": 2.007138363406702e-05,
      "loss": 0.0241,
      "step": 1819
    },
    {
      "epoch": 28.0,
      "grad_norm": 0.6328125,
      "learning_rate": 2.0064438128330975e-05,
      "loss": 0.0177,
      "step": 1820
    },
    {
      "epoch": 28.015384615384615,
      "grad_norm": 2.8125,
      "learning_rate": 2.005748893577969e-05,
      "loss": 0.0343,
      "step": 1821
    },
    {
      "epoch": 28.03076923076923,
      "grad_norm": 3.296875,
      "learning_rate": 2.0050536059800098e-05,
      "loss": 0.0315,
      "step": 1822
    },
    {
      "epoch": 28.046153846153846,
      "grad_norm": 1.8671875,
      "learning_rate": 2.0043579503780958e-05,
      "loss": 0.0261,
      "step": 1823
    },
    {
      "epoch": 28.06153846153846,
      "grad_norm": 1.7265625,
      "learning_rate": 2.003661927111281e-05,
      "loss": 0.039,
      "step": 1824
    },
    {
      "epoch": 28.076923076923077,
      "grad_norm": 3.953125,
      "learning_rate": 2.002965536518798e-05,
      "loss": 0.0341,
      "step": 1825
    },
    {
      "epoch": 28.092307692307692,
      "grad_norm": 2.046875,
      "learning_rate": 2.0022687789400607e-05,
      "loss": 0.0352,
      "step": 1826
    },
    {
      "epoch": 28.107692307692307,
      "grad_norm": 2.40625,
      "learning_rate": 2.0015716547146586e-05,
      "loss": 0.0357,
      "step": 1827
    },
    {
      "epoch": 28.123076923076923,
      "grad_norm": 2.21875,
      "learning_rate": 2.000874164182362e-05,
      "loss": 0.0226,
      "step": 1828
    },
    {
      "epoch": 28.138461538461538,
      "grad_norm": 1.25,
      "learning_rate": 2.0001763076831195e-05,
      "loss": 0.0555,
      "step": 1829
    },
    {
      "epoch": 28.153846153846153,
      "grad_norm": 1.9453125,
      "learning_rate": 1.999478085557058e-05,
      "loss": 0.0246,
      "step": 1830
    },
    {
      "epoch": 28.16923076923077,
      "grad_norm": 3.96875,
      "learning_rate": 1.9987794981444823e-05,
      "loss": 0.0375,
      "step": 1831
    },
    {
      "epoch": 28.184615384615384,
      "grad_norm": 0.482421875,
      "learning_rate": 1.9980805457858758e-05,
      "loss": 0.0396,
      "step": 1832
    },
    {
      "epoch": 28.2,
      "grad_norm": 1.15625,
      "learning_rate": 1.9973812288218987e-05,
      "loss": 0.0083,
      "step": 1833
    },
    {
      "epoch": 28.215384615384615,
      "grad_norm": 1.125,
      "learning_rate": 1.9966815475933902e-05,
      "loss": 0.0118,
      "step": 1834
    },
    {
      "epoch": 28.23076923076923,
      "grad_norm": 3.40625,
      "learning_rate": 1.9959815024413657e-05,
      "loss": 0.0587,
      "step": 1835
    },
    {
      "epoch": 28.246153846153845,
      "grad_norm": 0.515625,
      "learning_rate": 1.99528109370702e-05,
      "loss": 0.0055,
      "step": 1836
    },
    {
      "epoch": 28.26153846153846,
      "grad_norm": 1.0703125,
      "learning_rate": 1.9945803217317226e-05,
      "loss": 0.0127,
      "step": 1837
    },
    {
      "epoch": 28.276923076923076,
      "grad_norm": 1.390625,
      "learning_rate": 1.993879186857022e-05,
      "loss": 0.0299,
      "step": 1838
    },
    {
      "epoch": 28.29230769230769,
      "grad_norm": 0.279296875,
      "learning_rate": 1.9931776894246417e-05,
      "loss": 0.0246,
      "step": 1839
    },
    {
      "epoch": 28.307692307692307,
      "grad_norm": 0.61328125,
      "learning_rate": 1.992475829776484e-05,
      "loss": 0.0091,
      "step": 1840
    },
    {
      "epoch": 28.323076923076922,
      "grad_norm": 2.859375,
      "learning_rate": 1.991773608254627e-05,
      "loss": 0.0145,
      "step": 1841
    },
    {
      "epoch": 28.338461538461537,
      "grad_norm": 1.8125,
      "learning_rate": 1.9910710252013244e-05,
      "loss": 0.0145,
      "step": 1842
    },
    {
      "epoch": 28.353846153846153,
      "grad_norm": 2.75,
      "learning_rate": 1.9903680809590065e-05,
      "loss": 0.0227,
      "step": 1843
    },
    {
      "epoch": 28.369230769230768,
      "grad_norm": 3.28125,
      "learning_rate": 1.98966477587028e-05,
      "loss": 0.0174,
      "step": 1844
    },
    {
      "epoch": 28.384615384615383,
      "grad_norm": 3.921875,
      "learning_rate": 1.9889611102779276e-05,
      "loss": 0.0455,
      "step": 1845
    },
    {
      "epoch": 28.4,
      "grad_norm": 1.0546875,
      "learning_rate": 1.988257084524907e-05,
      "loss": 0.0157,
      "step": 1846
    },
    {
      "epoch": 28.415384615384614,
      "grad_norm": 2.484375,
      "learning_rate": 1.9875526989543517e-05,
      "loss": 0.0207,
      "step": 1847
    },
    {
      "epoch": 28.43076923076923,
      "grad_norm": 0.466796875,
      "learning_rate": 1.986847953909571e-05,
      "loss": 0.0064,
      "step": 1848
    },
    {
      "epoch": 28.446153846153845,
      "grad_norm": 1.2578125,
      "learning_rate": 1.9861428497340483e-05,
      "loss": 0.017,
      "step": 1849
    },
    {
      "epoch": 28.46153846153846,
      "grad_norm": 3.046875,
      "learning_rate": 1.9854373867714443e-05,
      "loss": 0.0319,
      "step": 1850
    },
    {
      "epoch": 28.476923076923075,
      "grad_norm": 1.734375,
      "learning_rate": 1.9847315653655915e-05,
      "loss": 0.0382,
      "step": 1851
    },
    {
      "epoch": 28.49230769230769,
      "grad_norm": 1.8515625,
      "learning_rate": 1.9840253858604995e-05,
      "loss": 0.0141,
      "step": 1852
    },
    {
      "epoch": 28.50769230769231,
      "grad_norm": 0.345703125,
      "learning_rate": 1.9833188486003516e-05,
      "loss": 0.0236,
      "step": 1853
    },
    {
      "epoch": 28.523076923076925,
      "grad_norm": 3.53125,
      "learning_rate": 1.9826119539295054e-05,
      "loss": 0.0385,
      "step": 1854
    },
    {
      "epoch": 28.53846153846154,
      "grad_norm": 1.5,
      "learning_rate": 1.9819047021924926e-05,
      "loss": 0.0166,
      "step": 1855
    },
    {
      "epoch": 28.553846153846155,
      "grad_norm": 0.1796875,
      "learning_rate": 1.9811970937340196e-05,
      "loss": 0.0178,
      "step": 1856
    },
    {
      "epoch": 28.56923076923077,
      "grad_norm": 4.125,
      "learning_rate": 1.9804891288989653e-05,
      "loss": 0.0299,
      "step": 1857
    },
    {
      "epoch": 28.584615384615386,
      "grad_norm": 1.3984375,
      "learning_rate": 1.9797808080323837e-05,
      "loss": 0.0628,
      "step": 1858
    },
    {
      "epoch": 28.6,
      "grad_norm": 0.80078125,
      "learning_rate": 1.979072131479502e-05,
      "loss": 0.0161,
      "step": 1859
    },
    {
      "epoch": 28.615384615384617,
      "grad_norm": 1.65625,
      "learning_rate": 1.9783630995857202e-05,
      "loss": 0.0076,
      "step": 1860
    },
    {
      "epoch": 28.630769230769232,
      "grad_norm": 0.80859375,
      "learning_rate": 1.977653712696612e-05,
      "loss": 0.0337,
      "step": 1861
    },
    {
      "epoch": 28.646153846153847,
      "grad_norm": 3.375,
      "learning_rate": 1.9769439711579234e-05,
      "loss": 0.0305,
      "step": 1862
    },
    {
      "epoch": 28.661538461538463,
      "grad_norm": 2.5,
      "learning_rate": 1.976233875315575e-05,
      "loss": 0.0198,
      "step": 1863
    },
    {
      "epoch": 28.676923076923078,
      "grad_norm": 0.6875,
      "learning_rate": 1.9755234255156576e-05,
      "loss": 0.023,
      "step": 1864
    },
    {
      "epoch": 28.692307692307693,
      "grad_norm": 0.65625,
      "learning_rate": 1.9748126221044367e-05,
      "loss": 0.0146,
      "step": 1865
    },
    {
      "epoch": 28.70769230769231,
      "grad_norm": 2.9375,
      "learning_rate": 1.9741014654283486e-05,
      "loss": 0.0472,
      "step": 1866
    },
    {
      "epoch": 28.723076923076924,
      "grad_norm": 1.6796875,
      "learning_rate": 1.973389955834003e-05,
      "loss": 0.0289,
      "step": 1867
    },
    {
      "epoch": 28.73846153846154,
      "grad_norm": 0.24609375,
      "learning_rate": 1.9726780936681807e-05,
      "loss": 0.0088,
      "step": 1868
    },
    {
      "epoch": 28.753846153846155,
      "grad_norm": 2.671875,
      "learning_rate": 1.9719658792778344e-05,
      "loss": 0.0412,
      "step": 1869
    },
    {
      "epoch": 28.76923076923077,
      "grad_norm": 0.62109375,
      "learning_rate": 1.9712533130100898e-05,
      "loss": 0.0188,
      "step": 1870
    },
    {
      "epoch": 28.784615384615385,
      "grad_norm": 2.5625,
      "learning_rate": 1.970540395212242e-05,
      "loss": 0.0448,
      "step": 1871
    },
    {
      "epoch": 28.8,
      "grad_norm": 0.412109375,
      "learning_rate": 1.9698271262317587e-05,
      "loss": 0.0041,
      "step": 1872
    },
    {
      "epoch": 28.815384615384616,
      "grad_norm": 0.6015625,
      "learning_rate": 1.9691135064162784e-05,
      "loss": 0.0091,
      "step": 1873
    },
    {
      "epoch": 28.83076923076923,
      "grad_norm": 6.8125,
      "learning_rate": 1.968399536113611e-05,
      "loss": 0.0415,
      "step": 1874
    },
    {
      "epoch": 28.846153846153847,
      "grad_norm": 0.203125,
      "learning_rate": 1.9676852156717375e-05,
      "loss": 0.0385,
      "step": 1875
    },
    {
      "epoch": 28.861538461538462,
      "grad_norm": 4.4375,
      "learning_rate": 1.9669705454388076e-05,
      "loss": 0.0291,
      "step": 1876
    },
    {
      "epoch": 28.876923076923077,
      "grad_norm": 1.953125,
      "learning_rate": 1.9662555257631445e-05,
      "loss": 0.0175,
      "step": 1877
    },
    {
      "epoch": 28.892307692307693,
      "grad_norm": 6.25,
      "learning_rate": 1.9655401569932388e-05,
      "loss": 0.0518,
      "step": 1878
    },
    {
      "epoch": 28.907692307692308,
      "grad_norm": 1.390625,
      "learning_rate": 1.9648244394777533e-05,
      "loss": 0.0289,
      "step": 1879
    },
    {
      "epoch": 28.923076923076923,
      "grad_norm": 7.34375,
      "learning_rate": 1.9641083735655193e-05,
      "loss": 0.0446,
      "step": 1880
    },
    {
      "epoch": 28.93846153846154,
      "grad_norm": 1.375,
      "learning_rate": 1.96339195960554e-05,
      "loss": 0.0135,
      "step": 1881
    },
    {
      "epoch": 28.953846153846154,
      "grad_norm": 4.03125,
      "learning_rate": 1.9626751979469855e-05,
      "loss": 0.0282,
      "step": 1882
    },
    {
      "epoch": 28.96923076923077,
      "grad_norm": 2.25,
      "learning_rate": 1.961958088939197e-05,
      "loss": 0.0129,
      "step": 1883
    },
    {
      "epoch": 28.984615384615385,
      "grad_norm": 1.421875,
      "learning_rate": 1.961240632931685e-05,
      "loss": 0.0167,
      "step": 1884
    },
    {
      "epoch": 29.0,
      "grad_norm": 0.19140625,
      "learning_rate": 1.9605228302741288e-05,
      "loss": 0.0075,
      "step": 1885
    },
    {
      "epoch": 29.015384615384615,
      "grad_norm": 4.1875,
      "learning_rate": 1.9598046813163766e-05,
      "loss": 0.0219,
      "step": 1886
    },
    {
      "epoch": 29.03076923076923,
      "grad_norm": 5.9375,
      "learning_rate": 1.9590861864084458e-05,
      "loss": 0.0483,
      "step": 1887
    },
    {
      "epoch": 29.046153846153846,
      "grad_norm": 0.796875,
      "learning_rate": 1.9583673459005212e-05,
      "loss": 0.0088,
      "step": 1888
    },
    {
      "epoch": 29.06153846153846,
      "grad_norm": 3.75,
      "learning_rate": 1.9576481601429584e-05,
      "loss": 0.0251,
      "step": 1889
    },
    {
      "epoch": 29.076923076923077,
      "grad_norm": 2.546875,
      "learning_rate": 1.9569286294862788e-05,
      "loss": 0.0258,
      "step": 1890
    },
    {
      "epoch": 29.092307692307692,
      "grad_norm": 2.125,
      "learning_rate": 1.9562087542811725e-05,
      "loss": 0.017,
      "step": 1891
    },
    {
      "epoch": 29.107692307692307,
      "grad_norm": 2.953125,
      "learning_rate": 1.955488534878499e-05,
      "loss": 0.0153,
      "step": 1892
    },
    {
      "epoch": 29.123076923076923,
      "grad_norm": 0.296875,
      "learning_rate": 1.9547679716292834e-05,
      "loss": 0.0167,
      "step": 1893
    },
    {
      "epoch": 29.138461538461538,
      "grad_norm": 1.2265625,
      "learning_rate": 1.954047064884721e-05,
      "loss": 0.0169,
      "step": 1894
    },
    {
      "epoch": 29.153846153846153,
      "grad_norm": 0.58984375,
      "learning_rate": 1.953325814996171e-05,
      "loss": 0.003,
      "step": 1895
    },
    {
      "epoch": 29.16923076923077,
      "grad_norm": 7.03125,
      "learning_rate": 1.9526042223151634e-05,
      "loss": 0.0448,
      "step": 1896
    },
    {
      "epoch": 29.184615384615384,
      "grad_norm": 7.84375,
      "learning_rate": 1.951882287193393e-05,
      "loss": 0.0229,
      "step": 1897
    },
    {
      "epoch": 29.2,
      "grad_norm": 1.53125,
      "learning_rate": 1.9511600099827225e-05,
      "loss": 0.0152,
      "step": 1898
    },
    {
      "epoch": 29.215384615384615,
      "grad_norm": 0.6484375,
      "learning_rate": 1.950437391035181e-05,
      "loss": 0.006,
      "step": 1899
    },
    {
      "epoch": 29.23076923076923,
      "grad_norm": 0.49609375,
      "learning_rate": 1.949714430702964e-05,
      "loss": 0.003,
      "step": 1900
    },
    {
      "epoch": 29.246153846153845,
      "grad_norm": 7.84375,
      "learning_rate": 1.9489911293384337e-05,
      "loss": 0.0295,
      "step": 1901
    },
    {
      "epoch": 29.26153846153846,
      "grad_norm": 0.2431640625,
      "learning_rate": 1.9482674872941183e-05,
      "loss": 0.0284,
      "step": 1902
    },
    {
      "epoch": 29.276923076923076,
      "grad_norm": 2.546875,
      "learning_rate": 1.9475435049227125e-05,
      "loss": 0.022,
      "step": 1903
    },
    {
      "epoch": 29.29230769230769,
      "grad_norm": 0.85546875,
      "learning_rate": 1.946819182577076e-05,
      "loss": 0.0491,
      "step": 1904
    },
    {
      "epoch": 29.307692307692307,
      "grad_norm": 1.5234375,
      "learning_rate": 1.9460945206102352e-05,
      "loss": 0.008,
      "step": 1905
    },
    {
      "epoch": 29.323076923076922,
      "grad_norm": 0.41796875,
      "learning_rate": 1.9453695193753812e-05,
      "loss": 0.0848,
      "step": 1906
    },
    {
      "epoch": 29.338461538461537,
      "grad_norm": 1.8828125,
      "learning_rate": 1.9446441792258707e-05,
      "loss": 0.0104,
      "step": 1907
    },
    {
      "epoch": 29.353846153846153,
      "grad_norm": 0.640625,
      "learning_rate": 1.943918500515226e-05,
      "loss": 0.0101,
      "step": 1908
    },
    {
      "epoch": 29.369230769230768,
      "grad_norm": 0.8359375,
      "learning_rate": 1.9431924835971342e-05,
      "loss": 0.0438,
      "step": 1909
    },
    {
      "epoch": 29.384615384615383,
      "grad_norm": 0.2275390625,
      "learning_rate": 1.9424661288254467e-05,
      "loss": 0.0195,
      "step": 1910
    },
    {
      "epoch": 29.4,
      "grad_norm": 0.5625,
      "learning_rate": 1.9417394365541803e-05,
      "loss": 0.0312,
      "step": 1911
    },
    {
      "epoch": 29.415384615384614,
      "grad_norm": 1.0859375,
      "learning_rate": 1.9410124071375158e-05,
      "loss": 0.0097,
      "step": 1912
    },
    {
      "epoch": 29.43076923076923,
      "grad_norm": 2.8125,
      "learning_rate": 1.9402850409297986e-05,
      "loss": 0.0201,
      "step": 1913
    },
    {
      "epoch": 29.446153846153845,
      "grad_norm": 2.625,
      "learning_rate": 1.939557338285538e-05,
      "loss": 0.0551,
      "step": 1914
    },
    {
      "epoch": 29.46153846153846,
      "grad_norm": 1.828125,
      "learning_rate": 1.9388292995594076e-05,
      "loss": 0.0312,
      "step": 1915
    },
    {
      "epoch": 29.476923076923075,
      "grad_norm": 3.328125,
      "learning_rate": 1.9381009251062447e-05,
      "loss": 0.011,
      "step": 1916
    },
    {
      "epoch": 29.49230769230769,
      "grad_norm": 0.95703125,
      "learning_rate": 1.93737221528105e-05,
      "loss": 0.0125,
      "step": 1917
    },
    {
      "epoch": 29.50769230769231,
      "grad_norm": 0.1611328125,
      "learning_rate": 1.9366431704389874e-05,
      "loss": 0.0189,
      "step": 1918
    },
    {
      "epoch": 29.523076923076925,
      "grad_norm": 0.625,
      "learning_rate": 1.9359137909353857e-05,
      "loss": 0.02,
      "step": 1919
    },
    {
      "epoch": 29.53846153846154,
      "grad_norm": 8.1875,
      "learning_rate": 1.9351840771257345e-05,
      "loss": 0.073,
      "step": 1920
    },
    {
      "epoch": 29.553846153846155,
      "grad_norm": 1.546875,
      "learning_rate": 1.934454029365688e-05,
      "loss": 0.0105,
      "step": 1921
    },
    {
      "epoch": 29.56923076923077,
      "grad_norm": 6.59375,
      "learning_rate": 1.933723648011062e-05,
      "loss": 0.0651,
      "step": 1922
    },
    {
      "epoch": 29.584615384615386,
      "grad_norm": 1.8203125,
      "learning_rate": 1.9329929334178366e-05,
      "loss": 0.0383,
      "step": 1923
    },
    {
      "epoch": 29.6,
      "grad_norm": 0.2109375,
      "learning_rate": 1.9322618859421522e-05,
      "loss": 0.0396,
      "step": 1924
    },
    {
      "epoch": 29.615384615384617,
      "grad_norm": 2.078125,
      "learning_rate": 1.9315305059403127e-05,
      "loss": 0.0116,
      "step": 1925
    },
    {
      "epoch": 29.630769230769232,
      "grad_norm": 1.203125,
      "learning_rate": 1.930798793768784e-05,
      "loss": 0.0136,
      "step": 1926
    },
    {
      "epoch": 29.646153846153847,
      "grad_norm": 2.140625,
      "learning_rate": 1.9300667497841934e-05,
      "loss": 0.0278,
      "step": 1927
    },
    {
      "epoch": 29.661538461538463,
      "grad_norm": 6.5,
      "learning_rate": 1.929334374343331e-05,
      "loss": 0.0536,
      "step": 1928
    },
    {
      "epoch": 29.676923076923078,
      "grad_norm": 2.0,
      "learning_rate": 1.9286016678031472e-05,
      "loss": 0.0317,
      "step": 1929
    },
    {
      "epoch": 29.692307692307693,
      "grad_norm": 2.25,
      "learning_rate": 1.9278686305207537e-05,
      "loss": 0.0291,
      "step": 1930
    },
    {
      "epoch": 29.70769230769231,
      "grad_norm": 1.5,
      "learning_rate": 1.927135262853425e-05,
      "loss": 0.0108,
      "step": 1931
    },
    {
      "epoch": 29.723076923076924,
      "grad_norm": 0.302734375,
      "learning_rate": 1.9264015651585948e-05,
      "loss": 0.0204,
      "step": 1932
    },
    {
      "epoch": 29.73846153846154,
      "grad_norm": 1.46875,
      "learning_rate": 1.925667537793859e-05,
      "loss": 0.0221,
      "step": 1933
    },
    {
      "epoch": 29.753846153846155,
      "grad_norm": 1.5234375,
      "learning_rate": 1.9249331811169735e-05,
      "loss": 0.0151,
      "step": 1934
    },
    {
      "epoch": 29.76923076923077,
      "grad_norm": 0.546875,
      "learning_rate": 1.9241984954858543e-05,
      "loss": 0.0161,
      "step": 1935
    },
    {
      "epoch": 29.784615384615385,
      "grad_norm": 1.1796875,
      "learning_rate": 1.9234634812585788e-05,
      "loss": 0.0124,
      "step": 1936
    },
    {
      "epoch": 29.8,
      "grad_norm": 0.423828125,
      "learning_rate": 1.9227281387933842e-05,
      "loss": 0.0145,
      "step": 1937
    },
    {
      "epoch": 29.815384615384616,
      "grad_norm": 2.90625,
      "learning_rate": 1.921992468448667e-05,
      "loss": 0.0258,
      "step": 1938
    },
    {
      "epoch": 29.83076923076923,
      "grad_norm": 8.0625,
      "learning_rate": 1.921256470582984e-05,
      "loss": 0.0643,
      "step": 1939
    },
    {
      "epoch": 29.846153846153847,
      "grad_norm": 3.8125,
      "learning_rate": 1.9205201455550514e-05,
      "loss": 0.0213,
      "step": 1940
    },
    {
      "epoch": 29.861538461538462,
      "grad_norm": 7.9375,
      "learning_rate": 1.9197834937237457e-05,
      "loss": 0.0199,
      "step": 1941
    },
    {
      "epoch": 29.876923076923077,
      "grad_norm": 3.15625,
      "learning_rate": 1.9190465154481018e-05,
      "loss": 0.0263,
      "step": 1942
    },
    {
      "epoch": 29.892307692307693,
      "grad_norm": 6.96875,
      "learning_rate": 1.918309211087314e-05,
      "loss": 0.0644,
      "step": 1943
    },
    {
      "epoch": 29.907692307692308,
      "grad_norm": 0.69140625,
      "learning_rate": 1.917571581000736e-05,
      "loss": 0.0085,
      "step": 1944
    },
    {
      "epoch": 29.923076923076923,
      "grad_norm": 2.953125,
      "learning_rate": 1.9168336255478787e-05,
      "loss": 0.0194,
      "step": 1945
    },
    {
      "epoch": 29.93846153846154,
      "grad_norm": 2.046875,
      "learning_rate": 1.9160953450884125e-05,
      "loss": 0.0275,
      "step": 1946
    },
    {
      "epoch": 29.953846153846154,
      "grad_norm": 1.2734375,
      "learning_rate": 1.9153567399821674e-05,
      "loss": 0.0131,
      "step": 1947
    },
    {
      "epoch": 29.96923076923077,
      "grad_norm": 2.671875,
      "learning_rate": 1.91461781058913e-05,
      "loss": 0.0304,
      "step": 1948
    },
    {
      "epoch": 29.984615384615385,
      "grad_norm": 0.392578125,
      "learning_rate": 1.913878557269446e-05,
      "loss": 0.0057,
      "step": 1949
    },
    {
      "epoch": 30.0,
      "grad_norm": 1.484375,
      "learning_rate": 1.9131389803834177e-05,
      "loss": 0.0222,
      "step": 1950
    },
    {
      "epoch": 30.015384615384615,
      "grad_norm": 1.484375,
      "learning_rate": 1.912399080291506e-05,
      "loss": 0.0377,
      "step": 1951
    },
    {
      "epoch": 30.03076923076923,
      "grad_norm": 1.3671875,
      "learning_rate": 1.9116588573543305e-05,
      "loss": 0.0101,
      "step": 1952
    },
    {
      "epoch": 30.046153846153846,
      "grad_norm": 0.26171875,
      "learning_rate": 1.9109183119326644e-05,
      "loss": 0.0024,
      "step": 1953
    },
    {
      "epoch": 30.06153846153846,
      "grad_norm": 1.0078125,
      "learning_rate": 1.9101774443874432e-05,
      "loss": 0.0073,
      "step": 1954
    },
    {
      "epoch": 30.076923076923077,
      "grad_norm": 6.09375,
      "learning_rate": 1.9094362550797547e-05,
      "loss": 0.0306,
      "step": 1955
    },
    {
      "epoch": 30.092307692307692,
      "grad_norm": 3.796875,
      "learning_rate": 1.9086947443708472e-05,
      "loss": 0.0514,
      "step": 1956
    },
    {
      "epoch": 30.107692307692307,
      "grad_norm": 9.3125,
      "learning_rate": 1.9079529126221223e-05,
      "loss": 0.034,
      "step": 1957
    },
    {
      "epoch": 30.123076923076923,
      "grad_norm": 2.984375,
      "learning_rate": 1.9072107601951417e-05,
      "loss": 0.0286,
      "step": 1958
    },
    {
      "epoch": 30.138461538461538,
      "grad_norm": 0.57421875,
      "learning_rate": 1.90646828745162e-05,
      "loss": 0.0032,
      "step": 1959
    },
    {
      "epoch": 30.153846153846153,
      "grad_norm": 2.515625,
      "learning_rate": 1.9057254947534293e-05,
      "loss": 0.0212,
      "step": 1960
    },
    {
      "epoch": 30.16923076923077,
      "grad_norm": 3.34375,
      "learning_rate": 1.9049823824625993e-05,
      "loss": 0.0322,
      "step": 1961
    },
    {
      "epoch": 30.184615384615384,
      "grad_norm": 0.8671875,
      "learning_rate": 1.904238950941312e-05,
      "loss": 0.0089,
      "step": 1962
    },
    {
      "epoch": 30.2,
      "grad_norm": 2.03125,
      "learning_rate": 1.9034952005519082e-05,
      "loss": 0.0166,
      "step": 1963
    },
    {
      "epoch": 30.215384615384615,
      "grad_norm": 7.46875,
      "learning_rate": 1.9027511316568824e-05,
      "loss": 0.023,
      "step": 1964
    },
    {
      "epoch": 30.23076923076923,
      "grad_norm": 1.625,
      "learning_rate": 1.902006744618885e-05,
      "loss": 0.0132,
      "step": 1965
    },
    {
      "epoch": 30.246153846153845,
      "grad_norm": 0.35546875,
      "learning_rate": 1.901262039800721e-05,
      "loss": 0.0306,
      "step": 1966
    },
    {
      "epoch": 30.26153846153846,
      "grad_norm": 5.375,
      "learning_rate": 1.9005170175653498e-05,
      "loss": 0.0592,
      "step": 1967
    },
    {
      "epoch": 30.276923076923076,
      "grad_norm": 3.609375,
      "learning_rate": 1.8997716782758867e-05,
      "loss": 0.028,
      "step": 1968
    },
    {
      "epoch": 30.29230769230769,
      "grad_norm": 1.8125,
      "learning_rate": 1.899026022295602e-05,
      "loss": 0.0162,
      "step": 1969
    },
    {
      "epoch": 30.307692307692307,
      "grad_norm": 4.125,
      "learning_rate": 1.8982800499879178e-05,
      "loss": 0.0337,
      "step": 1970
    },
    {
      "epoch": 30.323076923076922,
      "grad_norm": 5.65625,
      "learning_rate": 1.897533761716413e-05,
      "loss": 0.0724,
      "step": 1971
    },
    {
      "epoch": 30.338461538461537,
      "grad_norm": 1.6015625,
      "learning_rate": 1.8967871578448187e-05,
      "loss": 0.0132,
      "step": 1972
    },
    {
      "epoch": 30.353846153846153,
      "grad_norm": 3.1875,
      "learning_rate": 1.8960402387370214e-05,
      "loss": 0.0265,
      "step": 1973
    },
    {
      "epoch": 30.369230769230768,
      "grad_norm": 0.7734375,
      "learning_rate": 1.8952930047570595e-05,
      "loss": 0.0086,
      "step": 1974
    },
    {
      "epoch": 30.384615384615383,
      "grad_norm": 5.375,
      "learning_rate": 1.8945454562691266e-05,
      "loss": 0.0484,
      "step": 1975
    },
    {
      "epoch": 30.4,
      "grad_norm": 1.953125,
      "learning_rate": 1.893797593637568e-05,
      "loss": 0.0166,
      "step": 1976
    },
    {
      "epoch": 30.415384615384614,
      "grad_norm": 6.09375,
      "learning_rate": 1.893049417226883e-05,
      "loss": 0.0279,
      "step": 1977
    },
    {
      "epoch": 30.43076923076923,
      "grad_norm": 0.953125,
      "learning_rate": 1.8923009274017245e-05,
      "loss": 0.0122,
      "step": 1978
    },
    {
      "epoch": 30.446153846153845,
      "grad_norm": 2.671875,
      "learning_rate": 1.8915521245268964e-05,
      "loss": 0.014,
      "step": 1979
    },
    {
      "epoch": 30.46153846153846,
      "grad_norm": 2.546875,
      "learning_rate": 1.8908030089673563e-05,
      "loss": 0.0146,
      "step": 1980
    },
    {
      "epoch": 30.476923076923075,
      "grad_norm": 5.625,
      "learning_rate": 1.8900535810882146e-05,
      "loss": 0.0218,
      "step": 1981
    },
    {
      "epoch": 30.49230769230769,
      "grad_norm": 1.90625,
      "learning_rate": 1.8893038412547327e-05,
      "loss": 0.0258,
      "step": 1982
    },
    {
      "epoch": 30.50769230769231,
      "grad_norm": 0.953125,
      "learning_rate": 1.888553789832325e-05,
      "loss": 0.0038,
      "step": 1983
    },
    {
      "epoch": 30.523076923076925,
      "grad_norm": 6.59375,
      "learning_rate": 1.8878034271865576e-05,
      "loss": 0.052,
      "step": 1984
    },
    {
      "epoch": 30.53846153846154,
      "grad_norm": 0.96875,
      "learning_rate": 1.8870527536831473e-05,
      "loss": 0.0135,
      "step": 1985
    },
    {
      "epoch": 30.553846153846155,
      "grad_norm": 1.140625,
      "learning_rate": 1.8863017696879642e-05,
      "loss": 0.0169,
      "step": 1986
    },
    {
      "epoch": 30.56923076923077,
      "grad_norm": 5.1875,
      "learning_rate": 1.8855504755670287e-05,
      "loss": 0.0221,
      "step": 1987
    },
    {
      "epoch": 30.584615384615386,
      "grad_norm": 3.0625,
      "learning_rate": 1.8847988716865115e-05,
      "loss": 0.0175,
      "step": 1988
    },
    {
      "epoch": 30.6,
      "grad_norm": 1.53125,
      "learning_rate": 1.8840469584127367e-05,
      "loss": 0.0282,
      "step": 1989
    },
    {
      "epoch": 30.615384615384617,
      "grad_norm": 9.0625,
      "learning_rate": 1.8832947361121763e-05,
      "loss": 0.0374,
      "step": 1990
    },
    {
      "epoch": 30.630769230769232,
      "grad_norm": 4.15625,
      "learning_rate": 1.8825422051514546e-05,
      "loss": 0.0352,
      "step": 1991
    },
    {
      "epoch": 30.646153846153847,
      "grad_norm": 10.3125,
      "learning_rate": 1.8817893658973475e-05,
      "loss": 0.0385,
      "step": 1992
    },
    {
      "epoch": 30.661538461538463,
      "grad_norm": 1.328125,
      "learning_rate": 1.8810362187167774e-05,
      "loss": 0.0316,
      "step": 1993
    },
    {
      "epoch": 30.676923076923078,
      "grad_norm": 1.5078125,
      "learning_rate": 1.8802827639768206e-05,
      "loss": 0.0234,
      "step": 1994
    },
    {
      "epoch": 30.692307692307693,
      "grad_norm": 2.359375,
      "learning_rate": 1.879529002044701e-05,
      "loss": 0.0202,
      "step": 1995
    },
    {
      "epoch": 30.70769230769231,
      "grad_norm": 2.328125,
      "learning_rate": 1.878774933287794e-05,
      "loss": 0.0561,
      "step": 1996
    },
    {
      "epoch": 30.723076923076924,
      "grad_norm": 4.15625,
      "learning_rate": 1.8780205580736228e-05,
      "loss": 0.0421,
      "step": 1997
    },
    {
      "epoch": 30.73846153846154,
      "grad_norm": 1.2421875,
      "learning_rate": 1.8772658767698607e-05,
      "loss": 0.009,
      "step": 1998
    },
    {
      "epoch": 30.753846153846155,
      "grad_norm": 1.1015625,
      "learning_rate": 1.8765108897443303e-05,
      "loss": 0.0081,
      "step": 1999
    },
    {
      "epoch": 30.76923076923077,
      "grad_norm": 0.66796875,
      "learning_rate": 1.8757555973650036e-05,
      "loss": 0.016,
      "step": 2000
    },
    {
      "epoch": 30.784615384615385,
      "grad_norm": 5.59375,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.0409,
      "step": 2001
    },
    {
      "epoch": 30.8,
      "grad_norm": 3.921875,
      "learning_rate": 1.8742440980175896e-05,
      "loss": 0.032,
      "step": 2002
    },
    {
      "epoch": 30.815384615384616,
      "grad_norm": 0.90234375,
      "learning_rate": 1.8734878917861885e-05,
      "loss": 0.0288,
      "step": 2003
    },
    {
      "epoch": 30.83076923076923,
      "grad_norm": 2.0625,
      "learning_rate": 1.8727313816743637e-05,
      "loss": 0.0342,
      "step": 2004
    },
    {
      "epoch": 30.846153846153847,
      "grad_norm": 3.25,
      "learning_rate": 1.8719745680508282e-05,
      "loss": 0.0234,
      "step": 2005
    },
    {
      "epoch": 30.861538461538462,
      "grad_norm": 4.03125,
      "learning_rate": 1.8712174512844445e-05,
      "loss": 0.0352,
      "step": 2006
    },
    {
      "epoch": 30.876923076923077,
      "grad_norm": 0.1953125,
      "learning_rate": 1.8704600317442213e-05,
      "loss": 0.0137,
      "step": 2007
    },
    {
      "epoch": 30.892307692307693,
      "grad_norm": 2.734375,
      "learning_rate": 1.869702309799316e-05,
      "loss": 0.0634,
      "step": 2008
    },
    {
      "epoch": 30.907692307692308,
      "grad_norm": 1.7421875,
      "learning_rate": 1.8689442858190332e-05,
      "loss": 0.033,
      "step": 2009
    },
    {
      "epoch": 30.923076923076923,
      "grad_norm": 2.28125,
      "learning_rate": 1.8681859601728248e-05,
      "loss": 0.0525,
      "step": 2010
    },
    {
      "epoch": 30.93846153846154,
      "grad_norm": 3.9375,
      "learning_rate": 1.8674273332302888e-05,
      "loss": 0.0285,
      "step": 2011
    },
    {
      "epoch": 30.953846153846154,
      "grad_norm": 2.515625,
      "learning_rate": 1.8666684053611712e-05,
      "loss": 0.0187,
      "step": 2012
    },
    {
      "epoch": 30.96923076923077,
      "grad_norm": 11.875,
      "learning_rate": 1.8659091769353644e-05,
      "loss": 0.0457,
      "step": 2013
    },
    {
      "epoch": 30.984615384615385,
      "grad_norm": 1.875,
      "learning_rate": 1.865149648322907e-05,
      "loss": 0.0055,
      "step": 2014
    },
    {
      "epoch": 31.0,
      "grad_norm": 3.703125,
      "learning_rate": 1.8643898198939837e-05,
      "loss": 0.0199,
      "step": 2015
    },
    {
      "epoch": 31.015384615384615,
      "grad_norm": 2.59375,
      "learning_rate": 1.863629692018926e-05,
      "loss": 0.0526,
      "step": 2016
    },
    {
      "epoch": 31.03076923076923,
      "grad_norm": 1.2578125,
      "learning_rate": 1.862869265068211e-05,
      "loss": 0.0114,
      "step": 2017
    },
    {
      "epoch": 31.046153846153846,
      "grad_norm": 1.9375,
      "learning_rate": 1.8621085394124616e-05,
      "loss": 0.0325,
      "step": 2018
    },
    {
      "epoch": 31.06153846153846,
      "grad_norm": 2.3125,
      "learning_rate": 1.8613475154224454e-05,
      "loss": 0.0307,
      "step": 2019
    },
    {
      "epoch": 31.076923076923077,
      "grad_norm": 3.078125,
      "learning_rate": 1.8605861934690775e-05,
      "loss": 0.0386,
      "step": 2020
    },
    {
      "epoch": 31.092307692307692,
      "grad_norm": 3.0,
      "learning_rate": 1.859824573923416e-05,
      "loss": 0.0214,
      "step": 2021
    },
    {
      "epoch": 31.107692307692307,
      "grad_norm": 1.75,
      "learning_rate": 1.8590626571566654e-05,
      "loss": 0.018,
      "step": 2022
    },
    {
      "epoch": 31.123076923076923,
      "grad_norm": 2.546875,
      "learning_rate": 1.858300443540175e-05,
      "loss": 0.0405,
      "step": 2023
    },
    {
      "epoch": 31.138461538461538,
      "grad_norm": 1.3984375,
      "learning_rate": 1.857537933445438e-05,
      "loss": 0.0157,
      "step": 2024
    },
    {
      "epoch": 31.153846153846153,
      "grad_norm": 3.703125,
      "learning_rate": 1.8567751272440925e-05,
      "loss": 0.0347,
      "step": 2025
    },
    {
      "epoch": 31.16923076923077,
      "grad_norm": 3.125,
      "learning_rate": 1.8560120253079215e-05,
      "loss": 0.0491,
      "step": 2026
    },
    {
      "epoch": 31.184615384615384,
      "grad_norm": 2.0,
      "learning_rate": 1.855248628008851e-05,
      "loss": 0.0106,
      "step": 2027
    },
    {
      "epoch": 31.2,
      "grad_norm": 1.421875,
      "learning_rate": 1.854484935718952e-05,
      "loss": 0.0184,
      "step": 2028
    },
    {
      "epoch": 31.215384615384615,
      "grad_norm": 1.875,
      "learning_rate": 1.8537209488104383e-05,
      "loss": 0.0174,
      "step": 2029
    },
    {
      "epoch": 31.23076923076923,
      "grad_norm": 1.359375,
      "learning_rate": 1.852956667655669e-05,
      "loss": 0.0124,
      "step": 2030
    },
    {
      "epoch": 31.246153846153845,
      "grad_norm": 2.828125,
      "learning_rate": 1.852192092627144e-05,
      "loss": 0.0143,
      "step": 2031
    },
    {
      "epoch": 31.26153846153846,
      "grad_norm": 2.5625,
      "learning_rate": 1.8514272240975093e-05,
      "loss": 0.0397,
      "step": 2032
    },
    {
      "epoch": 31.276923076923076,
      "grad_norm": 4.65625,
      "learning_rate": 1.850662062439551e-05,
      "loss": 0.046,
      "step": 2033
    },
    {
      "epoch": 31.29230769230769,
      "grad_norm": 5.125,
      "learning_rate": 1.849896608026201e-05,
      "loss": 0.0283,
      "step": 2034
    },
    {
      "epoch": 31.307692307692307,
      "grad_norm": 0.490234375,
      "learning_rate": 1.8491308612305314e-05,
      "loss": 0.0295,
      "step": 2035
    },
    {
      "epoch": 31.323076923076922,
      "grad_norm": 0.7109375,
      "learning_rate": 1.848364822425759e-05,
      "loss": 0.0079,
      "step": 2036
    },
    {
      "epoch": 31.338461538461537,
      "grad_norm": 1.6953125,
      "learning_rate": 1.8475984919852404e-05,
      "loss": 0.0124,
      "step": 2037
    },
    {
      "epoch": 31.353846153846153,
      "grad_norm": 3.875,
      "learning_rate": 1.8468318702824765e-05,
      "loss": 0.0368,
      "step": 2038
    },
    {
      "epoch": 31.369230769230768,
      "grad_norm": 0.1953125,
      "learning_rate": 1.84606495769111e-05,
      "loss": 0.0143,
      "step": 2039
    },
    {
      "epoch": 31.384615384615383,
      "grad_norm": 2.578125,
      "learning_rate": 1.845297754584924e-05,
      "loss": 0.0223,
      "step": 2040
    },
    {
      "epoch": 31.4,
      "grad_norm": 2.828125,
      "learning_rate": 1.8445302613378444e-05,
      "loss": 0.0126,
      "step": 2041
    },
    {
      "epoch": 31.415384615384614,
      "grad_norm": 7.25,
      "learning_rate": 1.8437624783239372e-05,
      "loss": 0.0434,
      "step": 2042
    },
    {
      "epoch": 31.43076923076923,
      "grad_norm": 4.28125,
      "learning_rate": 1.8429944059174116e-05,
      "loss": 0.0559,
      "step": 2043
    },
    {
      "epoch": 31.446153846153845,
      "grad_norm": 10.0,
      "learning_rate": 1.8422260444926165e-05,
      "loss": 0.031,
      "step": 2044
    },
    {
      "epoch": 31.46153846153846,
      "grad_norm": 2.1875,
      "learning_rate": 1.8414573944240414e-05,
      "loss": 0.0239,
      "step": 2045
    },
    {
      "epoch": 31.476923076923075,
      "grad_norm": 2.484375,
      "learning_rate": 1.840688456086318e-05,
      "loss": 0.023,
      "step": 2046
    },
    {
      "epoch": 31.49230769230769,
      "grad_norm": 3.25,
      "learning_rate": 1.8399192298542163e-05,
      "loss": 0.03,
      "step": 2047
    },
    {
      "epoch": 31.50769230769231,
      "grad_norm": 0.4296875,
      "learning_rate": 1.839149716102648e-05,
      "loss": 0.0157,
      "step": 2048
    },
    {
      "epoch": 31.523076923076925,
      "grad_norm": 0.5546875,
      "learning_rate": 1.8383799152066662e-05,
      "loss": 0.0242,
      "step": 2049
    },
    {
      "epoch": 31.53846153846154,
      "grad_norm": 0.48828125,
      "learning_rate": 1.8376098275414606e-05,
      "loss": 0.0183,
      "step": 2050
    },
    {
      "epoch": 31.553846153846155,
      "grad_norm": 1.6328125,
      "learning_rate": 1.8368394534823635e-05,
      "loss": 0.0055,
      "step": 2051
    },
    {
      "epoch": 31.56923076923077,
      "grad_norm": 3.484375,
      "learning_rate": 1.8360687934048462e-05,
      "loss": 0.0355,
      "step": 2052
    },
    {
      "epoch": 31.584615384615386,
      "grad_norm": 3.65625,
      "learning_rate": 1.8352978476845182e-05,
      "loss": 0.0455,
      "step": 2053
    },
    {
      "epoch": 31.6,
      "grad_norm": 1.9140625,
      "learning_rate": 1.83452661669713e-05,
      "loss": 0.0329,
      "step": 2054
    },
    {
      "epoch": 31.615384615384617,
      "grad_norm": 3.8125,
      "learning_rate": 1.8337551008185694e-05,
      "loss": 0.025,
      "step": 2055
    },
    {
      "epoch": 31.630769230769232,
      "grad_norm": 2.46875,
      "learning_rate": 1.832983300424864e-05,
      "loss": 0.0208,
      "step": 2056
    },
    {
      "epoch": 31.646153846153847,
      "grad_norm": 0.7578125,
      "learning_rate": 1.8322112158921804e-05,
      "loss": 0.0167,
      "step": 2057
    },
    {
      "epoch": 31.661538461538463,
      "grad_norm": 1.515625,
      "learning_rate": 1.831438847596823e-05,
      "loss": 0.006,
      "step": 2058
    },
    {
      "epoch": 31.676923076923078,
      "grad_norm": 1.234375,
      "learning_rate": 1.830666195915235e-05,
      "loss": 0.0464,
      "step": 2059
    },
    {
      "epoch": 31.692307692307693,
      "grad_norm": 1.015625,
      "learning_rate": 1.8298932612239964e-05,
      "loss": 0.0055,
      "step": 2060
    },
    {
      "epoch": 31.70769230769231,
      "grad_norm": 6.0625,
      "learning_rate": 1.8291200438998273e-05,
      "loss": 0.0511,
      "step": 2061
    },
    {
      "epoch": 31.723076923076924,
      "grad_norm": 4.21875,
      "learning_rate": 1.8283465443195837e-05,
      "loss": 0.0345,
      "step": 2062
    },
    {
      "epoch": 31.73846153846154,
      "grad_norm": 0.95703125,
      "learning_rate": 1.827572762860261e-05,
      "loss": 0.0188,
      "step": 2063
    },
    {
      "epoch": 31.753846153846155,
      "grad_norm": 3.21875,
      "learning_rate": 1.8267986998989896e-05,
      "loss": 0.0692,
      "step": 2064
    },
    {
      "epoch": 31.76923076923077,
      "grad_norm": 0.68359375,
      "learning_rate": 1.8260243558130388e-05,
      "loss": 0.0075,
      "step": 2065
    },
    {
      "epoch": 31.784615384615385,
      "grad_norm": 6.0625,
      "learning_rate": 1.8252497309798147e-05,
      "loss": 0.0571,
      "step": 2066
    },
    {
      "epoch": 31.8,
      "grad_norm": 2.15625,
      "learning_rate": 1.8244748257768602e-05,
      "loss": 0.0521,
      "step": 2067
    },
    {
      "epoch": 31.815384615384616,
      "grad_norm": 0.265625,
      "learning_rate": 1.823699640581854e-05,
      "loss": 0.0025,
      "step": 2068
    },
    {
      "epoch": 31.83076923076923,
      "grad_norm": 2.015625,
      "learning_rate": 1.8229241757726124e-05,
      "loss": 0.0188,
      "step": 2069
    },
    {
      "epoch": 31.846153846153847,
      "grad_norm": 1.1640625,
      "learning_rate": 1.8221484317270876e-05,
      "loss": 0.0188,
      "step": 2070
    },
    {
      "epoch": 31.861538461538462,
      "grad_norm": 2.640625,
      "learning_rate": 1.821372408823368e-05,
      "loss": 0.0199,
      "step": 2071
    },
    {
      "epoch": 31.876923076923077,
      "grad_norm": 1.5859375,
      "learning_rate": 1.8205961074396772e-05,
      "loss": 0.0256,
      "step": 2072
    },
    {
      "epoch": 31.892307692307693,
      "grad_norm": 3.703125,
      "learning_rate": 1.8198195279543754e-05,
      "loss": 0.0293,
      "step": 2073
    },
    {
      "epoch": 31.907692307692308,
      "grad_norm": 1.40625,
      "learning_rate": 1.8190426707459572e-05,
      "loss": 0.017,
      "step": 2074
    },
    {
      "epoch": 31.923076923076923,
      "grad_norm": 0.66796875,
      "learning_rate": 1.818265536193055e-05,
      "loss": 0.0112,
      "step": 2075
    },
    {
      "epoch": 31.93846153846154,
      "grad_norm": 1.65625,
      "learning_rate": 1.8174881246744335e-05,
      "loss": 0.0183,
      "step": 2076
    },
    {
      "epoch": 31.953846153846154,
      "grad_norm": 5.90625,
      "learning_rate": 1.816710436568994e-05,
      "loss": 0.0679,
      "step": 2077
    },
    {
      "epoch": 31.96923076923077,
      "grad_norm": 0.90625,
      "learning_rate": 1.8159324722557725e-05,
      "loss": 0.0408,
      "step": 2078
    },
    {
      "epoch": 31.984615384615385,
      "grad_norm": 4.03125,
      "learning_rate": 1.8151542321139385e-05,
      "loss": 0.0374,
      "step": 2079
    },
    {
      "epoch": 32.0,
      "grad_norm": 2.734375,
      "learning_rate": 1.814375716522798e-05,
      "loss": 0.0187,
      "step": 2080
    },
    {
      "epoch": 32.01538461538462,
      "grad_norm": 3.84375,
      "learning_rate": 1.8135969258617884e-05,
      "loss": 0.036,
      "step": 2081
    },
    {
      "epoch": 32.03076923076923,
      "grad_norm": 2.9375,
      "learning_rate": 1.8128178605104844e-05,
      "loss": 0.0378,
      "step": 2082
    },
    {
      "epoch": 32.04615384615385,
      "grad_norm": 2.6875,
      "learning_rate": 1.8120385208485924e-05,
      "loss": 0.0198,
      "step": 2083
    },
    {
      "epoch": 32.06153846153846,
      "grad_norm": 2.0,
      "learning_rate": 1.811258907255953e-05,
      "loss": 0.0231,
      "step": 2084
    },
    {
      "epoch": 32.07692307692308,
      "grad_norm": 5.375,
      "learning_rate": 1.8104790201125408e-05,
      "loss": 0.0444,
      "step": 2085
    },
    {
      "epoch": 32.09230769230769,
      "grad_norm": 2.609375,
      "learning_rate": 1.809698859798462e-05,
      "loss": 0.0184,
      "step": 2086
    },
    {
      "epoch": 32.10769230769231,
      "grad_norm": 0.97265625,
      "learning_rate": 1.808918426693959e-05,
      "loss": 0.0493,
      "step": 2087
    },
    {
      "epoch": 32.12307692307692,
      "grad_norm": 1.8984375,
      "learning_rate": 1.8081377211794042e-05,
      "loss": 0.0278,
      "step": 2088
    },
    {
      "epoch": 32.13846153846154,
      "grad_norm": 2.703125,
      "learning_rate": 1.8073567436353054e-05,
      "loss": 0.0248,
      "step": 2089
    },
    {
      "epoch": 32.15384615384615,
      "grad_norm": 2.953125,
      "learning_rate": 1.8065754944422996e-05,
      "loss": 0.0279,
      "step": 2090
    },
    {
      "epoch": 32.16923076923077,
      "grad_norm": 2.078125,
      "learning_rate": 1.8057939739811595e-05,
      "loss": 0.0158,
      "step": 2091
    },
    {
      "epoch": 32.184615384615384,
      "grad_norm": 0.7109375,
      "learning_rate": 1.8050121826327883e-05,
      "loss": 0.0356,
      "step": 2092
    },
    {
      "epoch": 32.2,
      "grad_norm": 3.0,
      "learning_rate": 1.8042301207782214e-05,
      "loss": 0.0343,
      "step": 2093
    },
    {
      "epoch": 32.215384615384615,
      "grad_norm": 3.15625,
      "learning_rate": 1.8034477887986274e-05,
      "loss": 0.0246,
      "step": 2094
    },
    {
      "epoch": 32.23076923076923,
      "grad_norm": 0.7109375,
      "learning_rate": 1.8026651870753035e-05,
      "loss": 0.0187,
      "step": 2095
    },
    {
      "epoch": 32.246153846153845,
      "grad_norm": 3.921875,
      "learning_rate": 1.8018823159896816e-05,
      "loss": 0.0418,
      "step": 2096
    },
    {
      "epoch": 32.261538461538464,
      "grad_norm": 0.92578125,
      "learning_rate": 1.8010991759233235e-05,
      "loss": 0.006,
      "step": 2097
    },
    {
      "epoch": 32.276923076923076,
      "grad_norm": 0.83984375,
      "learning_rate": 1.8003157672579217e-05,
      "loss": 0.0135,
      "step": 2098
    },
    {
      "epoch": 32.292307692307695,
      "grad_norm": 1.5546875,
      "learning_rate": 1.7995320903753e-05,
      "loss": 0.0362,
      "step": 2099
    },
    {
      "epoch": 32.30769230769231,
      "grad_norm": 1.6796875,
      "learning_rate": 1.7987481456574135e-05,
      "loss": 0.0252,
      "step": 2100
    },
    {
      "epoch": 32.323076923076925,
      "grad_norm": 2.828125,
      "learning_rate": 1.7979639334863467e-05,
      "loss": 0.0723,
      "step": 2101
    },
    {
      "epoch": 32.33846153846154,
      "grad_norm": 2.453125,
      "learning_rate": 1.797179454244316e-05,
      "loss": 0.0144,
      "step": 2102
    },
    {
      "epoch": 32.353846153846156,
      "grad_norm": 1.765625,
      "learning_rate": 1.796394708313666e-05,
      "loss": 0.0158,
      "step": 2103
    },
    {
      "epoch": 32.36923076923077,
      "grad_norm": 2.484375,
      "learning_rate": 1.7956096960768738e-05,
      "loss": 0.0145,
      "step": 2104
    },
    {
      "epoch": 32.38461538461539,
      "grad_norm": 0.89453125,
      "learning_rate": 1.7948244179165428e-05,
      "loss": 0.0189,
      "step": 2105
    },
    {
      "epoch": 32.4,
      "grad_norm": 0.625,
      "learning_rate": 1.7940388742154096e-05,
      "loss": 0.0214,
      "step": 2106
    },
    {
      "epoch": 32.41538461538462,
      "grad_norm": 9.375,
      "learning_rate": 1.793253065356338e-05,
      "loss": 0.0406,
      "step": 2107
    },
    {
      "epoch": 32.43076923076923,
      "grad_norm": 5.875,
      "learning_rate": 1.792466991722322e-05,
      "loss": 0.0322,
      "step": 2108
    },
    {
      "epoch": 32.44615384615385,
      "grad_norm": 1.0390625,
      "learning_rate": 1.7916806536964842e-05,
      "loss": 0.0171,
      "step": 2109
    },
    {
      "epoch": 32.46153846153846,
      "grad_norm": 1.671875,
      "learning_rate": 1.790894051662076e-05,
      "loss": 0.0104,
      "step": 2110
    },
    {
      "epoch": 32.47692307692308,
      "grad_norm": 1.0859375,
      "learning_rate": 1.790107186002478e-05,
      "loss": 0.0416,
      "step": 2111
    },
    {
      "epoch": 32.49230769230769,
      "grad_norm": 8.25,
      "learning_rate": 1.7893200571011988e-05,
      "loss": 0.074,
      "step": 2112
    },
    {
      "epoch": 32.50769230769231,
      "grad_norm": 2.09375,
      "learning_rate": 1.7885326653418757e-05,
      "loss": 0.0277,
      "step": 2113
    },
    {
      "epoch": 32.52307692307692,
      "grad_norm": 4.0625,
      "learning_rate": 1.7877450111082733e-05,
      "loss": 0.0313,
      "step": 2114
    },
    {
      "epoch": 32.53846153846154,
      "grad_norm": 0.765625,
      "learning_rate": 1.7869570947842855e-05,
      "loss": 0.0286,
      "step": 2115
    },
    {
      "epoch": 32.55384615384615,
      "grad_norm": 4.0625,
      "learning_rate": 1.7861689167539324e-05,
      "loss": 0.0364,
      "step": 2116
    },
    {
      "epoch": 32.56923076923077,
      "grad_norm": 0.67578125,
      "learning_rate": 1.785380477401363e-05,
      "loss": 0.0056,
      "step": 2117
    },
    {
      "epoch": 32.58461538461538,
      "grad_norm": 4.3125,
      "learning_rate": 1.7845917771108527e-05,
      "loss": 0.0448,
      "step": 2118
    },
    {
      "epoch": 32.6,
      "grad_norm": 4.40625,
      "learning_rate": 1.783802816266804e-05,
      "loss": 0.0275,
      "step": 2119
    },
    {
      "epoch": 32.61538461538461,
      "grad_norm": 2.78125,
      "learning_rate": 1.7830135952537487e-05,
      "loss": 0.0167,
      "step": 2120
    },
    {
      "epoch": 32.63076923076923,
      "grad_norm": 3.796875,
      "learning_rate": 1.782224114456341e-05,
      "loss": 0.0422,
      "step": 2121
    },
    {
      "epoch": 32.646153846153844,
      "grad_norm": 1.2890625,
      "learning_rate": 1.781434374259366e-05,
      "loss": 0.0453,
      "step": 2122
    },
    {
      "epoch": 32.66153846153846,
      "grad_norm": 0.50390625,
      "learning_rate": 1.7806443750477324e-05,
      "loss": 0.0329,
      "step": 2123
    },
    {
      "epoch": 32.676923076923075,
      "grad_norm": 0.55078125,
      "learning_rate": 1.779854117206477e-05,
      "loss": 0.0293,
      "step": 2124
    },
    {
      "epoch": 32.69230769230769,
      "grad_norm": 2.359375,
      "learning_rate": 1.7790636011207622e-05,
      "loss": 0.0191,
      "step": 2125
    },
    {
      "epoch": 32.707692307692305,
      "grad_norm": 3.515625,
      "learning_rate": 1.7782728271758746e-05,
      "loss": 0.0497,
      "step": 2126
    },
    {
      "epoch": 32.723076923076924,
      "grad_norm": 1.09375,
      "learning_rate": 1.7774817957572286e-05,
      "loss": 0.0232,
      "step": 2127
    },
    {
      "epoch": 32.738461538461536,
      "grad_norm": 2.171875,
      "learning_rate": 1.7766905072503636e-05,
      "loss": 0.0098,
      "step": 2128
    },
    {
      "epoch": 32.753846153846155,
      "grad_norm": 0.91796875,
      "learning_rate": 1.7758989620409435e-05,
      "loss": 0.0176,
      "step": 2129
    },
    {
      "epoch": 32.76923076923077,
      "grad_norm": 2.671875,
      "learning_rate": 1.7751071605147582e-05,
      "loss": 0.0241,
      "step": 2130
    },
    {
      "epoch": 32.784615384615385,
      "grad_norm": 2.859375,
      "learning_rate": 1.7743151030577214e-05,
      "loss": 0.021,
      "step": 2131
    },
    {
      "epoch": 32.8,
      "grad_norm": 2.75,
      "learning_rate": 1.773522790055873e-05,
      "loss": 0.0126,
      "step": 2132
    },
    {
      "epoch": 32.815384615384616,
      "grad_norm": 2.328125,
      "learning_rate": 1.7727302218953766e-05,
      "loss": 0.0483,
      "step": 2133
    },
    {
      "epoch": 32.83076923076923,
      "grad_norm": 9.625,
      "learning_rate": 1.771937398962521e-05,
      "loss": 0.0424,
      "step": 2134
    },
    {
      "epoch": 32.84615384615385,
      "grad_norm": 0.7265625,
      "learning_rate": 1.771144321643717e-05,
      "loss": 0.0285,
      "step": 2135
    },
    {
      "epoch": 32.86153846153846,
      "grad_norm": 0.1572265625,
      "learning_rate": 1.7703509903255015e-05,
      "loss": 0.0159,
      "step": 2136
    },
    {
      "epoch": 32.87692307692308,
      "grad_norm": 2.953125,
      "learning_rate": 1.769557405394535e-05,
      "loss": 0.0315,
      "step": 2137
    },
    {
      "epoch": 32.89230769230769,
      "grad_norm": 3.84375,
      "learning_rate": 1.7687635672376006e-05,
      "loss": 0.0202,
      "step": 2138
    },
    {
      "epoch": 32.90769230769231,
      "grad_norm": 1.6875,
      "learning_rate": 1.7679694762416056e-05,
      "loss": 0.032,
      "step": 2139
    },
    {
      "epoch": 32.92307692307692,
      "grad_norm": 0.97265625,
      "learning_rate": 1.7671751327935797e-05,
      "loss": 0.0076,
      "step": 2140
    },
    {
      "epoch": 32.93846153846154,
      "grad_norm": 1.8671875,
      "learning_rate": 1.7663805372806775e-05,
      "loss": 0.0448,
      "step": 2141
    },
    {
      "epoch": 32.95384615384615,
      "grad_norm": 6.1875,
      "learning_rate": 1.765585690090174e-05,
      "loss": 0.0513,
      "step": 2142
    },
    {
      "epoch": 32.96923076923077,
      "grad_norm": 5.8125,
      "learning_rate": 1.7647905916094685e-05,
      "loss": 0.0389,
      "step": 2143
    },
    {
      "epoch": 32.98461538461538,
      "grad_norm": 8.125,
      "learning_rate": 1.763995242226083e-05,
      "loss": 0.0412,
      "step": 2144
    },
    {
      "epoch": 33.0,
      "grad_norm": 1.609375,
      "learning_rate": 1.76319964232766e-05,
      "loss": 0.0141,
      "step": 2145
    },
    {
      "epoch": 33.01538461538462,
      "grad_norm": 0.3359375,
      "learning_rate": 1.762403792301966e-05,
      "loss": 0.0101,
      "step": 2146
    },
    {
      "epoch": 33.03076923076923,
      "grad_norm": 1.5390625,
      "learning_rate": 1.7616076925368885e-05,
      "loss": 0.0117,
      "step": 2147
    },
    {
      "epoch": 33.04615384615385,
      "grad_norm": 0.64453125,
      "learning_rate": 1.7608113434204366e-05,
      "loss": 0.0381,
      "step": 2148
    },
    {
      "epoch": 33.06153846153846,
      "grad_norm": 5.5,
      "learning_rate": 1.7600147453407412e-05,
      "loss": 0.029,
      "step": 2149
    },
    {
      "epoch": 33.07692307692308,
      "grad_norm": 2.234375,
      "learning_rate": 1.7592178986860544e-05,
      "loss": 0.016,
      "step": 2150
    },
    {
      "epoch": 33.09230769230769,
      "grad_norm": 1.0625,
      "learning_rate": 1.7584208038447505e-05,
      "loss": 0.008,
      "step": 2151
    },
    {
      "epoch": 33.10769230769231,
      "grad_norm": 0.89453125,
      "learning_rate": 1.7576234612053227e-05,
      "loss": 0.0225,
      "step": 2152
    },
    {
      "epoch": 33.12307692307692,
      "grad_norm": 0.4296875,
      "learning_rate": 1.756825871156387e-05,
      "loss": 0.0076,
      "step": 2153
    },
    {
      "epoch": 33.13846153846154,
      "grad_norm": 5.5,
      "learning_rate": 1.756028034086678e-05,
      "loss": 0.0279,
      "step": 2154
    },
    {
      "epoch": 33.15384615384615,
      "grad_norm": 3.03125,
      "learning_rate": 1.7552299503850527e-05,
      "loss": 0.0284,
      "step": 2155
    },
    {
      "epoch": 33.16923076923077,
      "grad_norm": 0.6796875,
      "learning_rate": 1.7544316204404877e-05,
      "loss": 0.0146,
      "step": 2156
    },
    {
      "epoch": 33.184615384615384,
      "grad_norm": 0.79296875,
      "learning_rate": 1.7536330446420785e-05,
      "loss": 0.0091,
      "step": 2157
    },
    {
      "epoch": 33.2,
      "grad_norm": 0.578125,
      "learning_rate": 1.7528342233790415e-05,
      "loss": 0.0338,
      "step": 2158
    },
    {
      "epoch": 33.215384615384615,
      "grad_norm": 1.5234375,
      "learning_rate": 1.7520351570407127e-05,
      "loss": 0.0431,
      "step": 2159
    },
    {
      "epoch": 33.23076923076923,
      "grad_norm": 1.9140625,
      "learning_rate": 1.751235846016547e-05,
      "loss": 0.016,
      "step": 2160
    },
    {
      "epoch": 33.246153846153845,
      "grad_norm": 0.1845703125,
      "learning_rate": 1.7504362906961193e-05,
      "loss": 0.0219,
      "step": 2161
    },
    {
      "epoch": 33.261538461538464,
      "grad_norm": 1.65625,
      "learning_rate": 1.749636491469122e-05,
      "loss": 0.0183,
      "step": 2162
    },
    {
      "epoch": 33.276923076923076,
      "grad_norm": 1.140625,
      "learning_rate": 1.7488364487253692e-05,
      "loss": 0.0316,
      "step": 2163
    },
    {
      "epoch": 33.292307692307695,
      "grad_norm": 8.6875,
      "learning_rate": 1.748036162854791e-05,
      "loss": 0.0612,
      "step": 2164
    },
    {
      "epoch": 33.30769230769231,
      "grad_norm": 2.1875,
      "learning_rate": 1.7472356342474376e-05,
      "loss": 0.0218,
      "step": 2165
    },
    {
      "epoch": 33.323076923076925,
      "grad_norm": 1.3125,
      "learning_rate": 1.7464348632934758e-05,
      "loss": 0.0066,
      "step": 2166
    },
    {
      "epoch": 33.33846153846154,
      "grad_norm": 0.287109375,
      "learning_rate": 1.745633850383193e-05,
      "loss": 0.0105,
      "step": 2167
    },
    {
      "epoch": 33.353846153846156,
      "grad_norm": 2.65625,
      "learning_rate": 1.7448325959069923e-05,
      "loss": 0.012,
      "step": 2168
    },
    {
      "epoch": 33.36923076923077,
      "grad_norm": 0.703125,
      "learning_rate": 1.7440311002553957e-05,
      "loss": 0.0267,
      "step": 2169
    },
    {
      "epoch": 33.38461538461539,
      "grad_norm": 3.34375,
      "learning_rate": 1.7432293638190425e-05,
      "loss": 0.0215,
      "step": 2170
    },
    {
      "epoch": 33.4,
      "grad_norm": 0.51953125,
      "learning_rate": 1.742427386988689e-05,
      "loss": 0.0046,
      "step": 2171
    },
    {
      "epoch": 33.41538461538462,
      "grad_norm": 2.28125,
      "learning_rate": 1.741625170155209e-05,
      "loss": 0.0767,
      "step": 2172
    },
    {
      "epoch": 33.43076923076923,
      "grad_norm": 5.65625,
      "learning_rate": 1.740822713709594e-05,
      "loss": 0.0329,
      "step": 2173
    },
    {
      "epoch": 33.44615384615385,
      "grad_norm": 1.578125,
      "learning_rate": 1.7400200180429503e-05,
      "loss": 0.0135,
      "step": 2174
    },
    {
      "epoch": 33.46153846153846,
      "grad_norm": 5.625,
      "learning_rate": 1.7392170835465032e-05,
      "loss": 0.0505,
      "step": 2175
    },
    {
      "epoch": 33.47692307692308,
      "grad_norm": 1.4765625,
      "learning_rate": 1.738413910611592e-05,
      "loss": 0.0115,
      "step": 2176
    },
    {
      "epoch": 33.49230769230769,
      "grad_norm": 1.625,
      "learning_rate": 1.7376104996296746e-05,
      "loss": 0.0106,
      "step": 2177
    },
    {
      "epoch": 33.50769230769231,
      "grad_norm": 0.7109375,
      "learning_rate": 1.7368068509923233e-05,
      "loss": 0.0065,
      "step": 2178
    },
    {
      "epoch": 33.52307692307692,
      "grad_norm": 0.76171875,
      "learning_rate": 1.7360029650912263e-05,
      "loss": 0.0053,
      "step": 2179
    },
    {
      "epoch": 33.53846153846154,
      "grad_norm": 0.28515625,
      "learning_rate": 1.7351988423181885e-05,
      "loss": 0.0323,
      "step": 2180
    },
    {
      "epoch": 33.55384615384615,
      "grad_norm": 0.90625,
      "learning_rate": 1.7343944830651288e-05,
      "loss": 0.0141,
      "step": 2181
    },
    {
      "epoch": 33.56923076923077,
      "grad_norm": 6.59375,
      "learning_rate": 1.733589887724083e-05,
      "loss": 0.0219,
      "step": 2182
    },
    {
      "epoch": 33.58461538461538,
      "grad_norm": 1.34375,
      "learning_rate": 1.7327850566872e-05,
      "loss": 0.0217,
      "step": 2183
    },
    {
      "epoch": 33.6,
      "grad_norm": 0.5625,
      "learning_rate": 1.731979990346746e-05,
      "loss": 0.0347,
      "step": 2184
    },
    {
      "epoch": 33.61538461538461,
      "grad_norm": 2.921875,
      "learning_rate": 1.7311746890950994e-05,
      "loss": 0.0171,
      "step": 2185
    },
    {
      "epoch": 33.63076923076923,
      "grad_norm": 4.4375,
      "learning_rate": 1.7303691533247546e-05,
      "loss": 0.019,
      "step": 2186
    },
    {
      "epoch": 33.646153846153844,
      "grad_norm": 2.59375,
      "learning_rate": 1.7295633834283205e-05,
      "loss": 0.0116,
      "step": 2187
    },
    {
      "epoch": 33.66153846153846,
      "grad_norm": 3.328125,
      "learning_rate": 1.728757379798519e-05,
      "loss": 0.0192,
      "step": 2188
    },
    {
      "epoch": 33.676923076923075,
      "grad_norm": 5.21875,
      "learning_rate": 1.7279511428281864e-05,
      "loss": 0.0133,
      "step": 2189
    },
    {
      "epoch": 33.69230769230769,
      "grad_norm": 0.66796875,
      "learning_rate": 1.727144672910273e-05,
      "loss": 0.0242,
      "step": 2190
    },
    {
      "epoch": 33.707692307692305,
      "grad_norm": 0.546875,
      "learning_rate": 1.726337970437843e-05,
      "loss": 0.0056,
      "step": 2191
    },
    {
      "epoch": 33.723076923076924,
      "grad_norm": 0.9765625,
      "learning_rate": 1.7255310358040723e-05,
      "loss": 0.042,
      "step": 2192
    },
    {
      "epoch": 33.738461538461536,
      "grad_norm": 1.6875,
      "learning_rate": 1.7247238694022517e-05,
      "loss": 0.0051,
      "step": 2193
    },
    {
      "epoch": 33.753846153846155,
      "grad_norm": 0.74609375,
      "learning_rate": 1.7239164716257837e-05,
      "loss": 0.0467,
      "step": 2194
    },
    {
      "epoch": 33.76923076923077,
      "grad_norm": 0.72265625,
      "learning_rate": 1.7231088428681852e-05,
      "loss": 0.0326,
      "step": 2195
    },
    {
      "epoch": 33.784615384615385,
      "grad_norm": 12.5625,
      "learning_rate": 1.722300983523084e-05,
      "loss": 0.0565,
      "step": 2196
    },
    {
      "epoch": 33.8,
      "grad_norm": 3.625,
      "learning_rate": 1.721492893984221e-05,
      "loss": 0.0194,
      "step": 2197
    },
    {
      "epoch": 33.815384615384616,
      "grad_norm": 0.88671875,
      "learning_rate": 1.7206845746454492e-05,
      "loss": 0.0331,
      "step": 2198
    },
    {
      "epoch": 33.83076923076923,
      "grad_norm": 8.3125,
      "learning_rate": 1.7198760259007336e-05,
      "loss": 0.0402,
      "step": 2199
    },
    {
      "epoch": 33.84615384615385,
      "grad_norm": 0.2099609375,
      "learning_rate": 1.719067248144151e-05,
      "loss": 0.0233,
      "step": 2200
    },
    {
      "epoch": 33.86153846153846,
      "grad_norm": 0.4921875,
      "learning_rate": 1.7182582417698903e-05,
      "loss": 0.0093,
      "step": 2201
    },
    {
      "epoch": 33.87692307692308,
      "grad_norm": 5.1875,
      "learning_rate": 1.7174490071722504e-05,
      "loss": 0.0447,
      "step": 2202
    },
    {
      "epoch": 33.89230769230769,
      "grad_norm": 2.65625,
      "learning_rate": 1.716639544745644e-05,
      "loss": 0.0197,
      "step": 2203
    },
    {
      "epoch": 33.90769230769231,
      "grad_norm": 6.0625,
      "learning_rate": 1.7158298548845914e-05,
      "loss": 0.0662,
      "step": 2204
    },
    {
      "epoch": 33.92307692307692,
      "grad_norm": 1.1875,
      "learning_rate": 1.715019937983727e-05,
      "loss": 0.0224,
      "step": 2205
    },
    {
      "epoch": 33.93846153846154,
      "grad_norm": 1.3046875,
      "learning_rate": 1.714209794437794e-05,
      "loss": 0.0148,
      "step": 2206
    },
    {
      "epoch": 33.95384615384615,
      "grad_norm": 2.40625,
      "learning_rate": 1.7133994246416456e-05,
      "loss": 0.0156,
      "step": 2207
    },
    {
      "epoch": 33.96923076923077,
      "grad_norm": 3.875,
      "learning_rate": 1.712588828990248e-05,
      "loss": 0.0713,
      "step": 2208
    },
    {
      "epoch": 33.98461538461538,
      "grad_norm": 4.75,
      "learning_rate": 1.7117780078786745e-05,
      "loss": 0.0396,
      "step": 2209
    },
    {
      "epoch": 34.0,
      "grad_norm": 1.9765625,
      "learning_rate": 1.7109669617021095e-05,
      "loss": 0.0119,
      "step": 2210
    },
    {
      "epoch": 34.01538461538462,
      "grad_norm": 5.9375,
      "learning_rate": 1.7101556908558478e-05,
      "loss": 0.051,
      "step": 2211
    },
    {
      "epoch": 34.03076923076923,
      "grad_norm": 1.390625,
      "learning_rate": 1.7093441957352917e-05,
      "loss": 0.0102,
      "step": 2212
    },
    {
      "epoch": 34.04615384615385,
      "grad_norm": 0.77734375,
      "learning_rate": 1.7085324767359554e-05,
      "loss": 0.0078,
      "step": 2213
    },
    {
      "epoch": 34.06153846153846,
      "grad_norm": 2.96875,
      "learning_rate": 1.7077205342534598e-05,
      "loss": 0.0264,
      "step": 2214
    },
    {
      "epoch": 34.07692307692308,
      "grad_norm": 0.2099609375,
      "learning_rate": 1.7069083686835363e-05,
      "loss": 0.0036,
      "step": 2215
    },
    {
      "epoch": 34.09230769230769,
      "grad_norm": 1.2734375,
      "learning_rate": 1.7060959804220246e-05,
      "loss": 0.0154,
      "step": 2216
    },
    {
      "epoch": 34.10769230769231,
      "grad_norm": 1.609375,
      "learning_rate": 1.705283369864873e-05,
      "loss": 0.0155,
      "step": 2217
    },
    {
      "epoch": 34.12307692307692,
      "grad_norm": 2.765625,
      "learning_rate": 1.7044705374081374e-05,
      "loss": 0.0634,
      "step": 2218
    },
    {
      "epoch": 34.13846153846154,
      "grad_norm": 1.2578125,
      "learning_rate": 1.703657483447983e-05,
      "loss": 0.0157,
      "step": 2219
    },
    {
      "epoch": 34.15384615384615,
      "grad_norm": 0.431640625,
      "learning_rate": 1.702844208380682e-05,
      "loss": 0.0124,
      "step": 2220
    },
    {
      "epoch": 34.16923076923077,
      "grad_norm": 6.03125,
      "learning_rate": 1.7020307126026152e-05,
      "loss": 0.0464,
      "step": 2221
    },
    {
      "epoch": 34.184615384615384,
      "grad_norm": 0.54296875,
      "learning_rate": 1.7012169965102708e-05,
      "loss": 0.0131,
      "step": 2222
    },
    {
      "epoch": 34.2,
      "grad_norm": 4.3125,
      "learning_rate": 1.7004030605002434e-05,
      "loss": 0.0325,
      "step": 2223
    },
    {
      "epoch": 34.215384615384615,
      "grad_norm": 0.66015625,
      "learning_rate": 1.699588904969236e-05,
      "loss": 0.033,
      "step": 2224
    },
    {
      "epoch": 34.23076923076923,
      "grad_norm": 2.109375,
      "learning_rate": 1.698774530314058e-05,
      "loss": 0.0103,
      "step": 2225
    },
    {
      "epoch": 34.246153846153845,
      "grad_norm": 2.765625,
      "learning_rate": 1.697959936931625e-05,
      "loss": 0.072,
      "step": 2226
    },
    {
      "epoch": 34.261538461538464,
      "grad_norm": 0.7734375,
      "learning_rate": 1.6971451252189617e-05,
      "loss": 0.0172,
      "step": 2227
    },
    {
      "epoch": 34.276923076923076,
      "grad_norm": 1.4453125,
      "learning_rate": 1.6963300955731958e-05,
      "loss": 0.0128,
      "step": 2228
    },
    {
      "epoch": 34.292307692307695,
      "grad_norm": 1.9375,
      "learning_rate": 1.6955148483915633e-05,
      "loss": 0.0277,
      "step": 2229
    },
    {
      "epoch": 34.30769230769231,
      "grad_norm": 3.1875,
      "learning_rate": 1.694699384071406e-05,
      "loss": 0.0188,
      "step": 2230
    },
    {
      "epoch": 34.323076923076925,
      "grad_norm": 1.0546875,
      "learning_rate": 1.6938837030101714e-05,
      "loss": 0.0219,
      "step": 2231
    },
    {
      "epoch": 34.33846153846154,
      "grad_norm": 4.28125,
      "learning_rate": 1.6930678056054123e-05,
      "loss": 0.0383,
      "step": 2232
    },
    {
      "epoch": 34.353846153846156,
      "grad_norm": 0.58984375,
      "learning_rate": 1.692251692254787e-05,
      "loss": 0.0125,
      "step": 2233
    },
    {
      "epoch": 34.36923076923077,
      "grad_norm": 1.0546875,
      "learning_rate": 1.6914353633560594e-05,
      "loss": 0.0259,
      "step": 2234
    },
    {
      "epoch": 34.38461538461539,
      "grad_norm": 3.234375,
      "learning_rate": 1.690618819307098e-05,
      "loss": 0.0125,
      "step": 2235
    },
    {
      "epoch": 34.4,
      "grad_norm": 0.56640625,
      "learning_rate": 1.6898020605058773e-05,
      "loss": 0.008,
      "step": 2236
    },
    {
      "epoch": 34.41538461538462,
      "grad_norm": 2.6875,
      "learning_rate": 1.6889850873504752e-05,
      "loss": 0.0203,
      "step": 2237
    },
    {
      "epoch": 34.43076923076923,
      "grad_norm": 0.7109375,
      "learning_rate": 1.6881679002390738e-05,
      "loss": 0.0164,
      "step": 2238
    },
    {
      "epoch": 34.44615384615385,
      "grad_norm": 0.86328125,
      "learning_rate": 1.6873504995699614e-05,
      "loss": 0.0073,
      "step": 2239
    },
    {
      "epoch": 34.46153846153846,
      "grad_norm": 2.46875,
      "learning_rate": 1.686532885741528e-05,
      "loss": 0.0188,
      "step": 2240
    },
    {
      "epoch": 34.47692307692308,
      "grad_norm": 1.1328125,
      "learning_rate": 1.6857150591522692e-05,
      "loss": 0.0348,
      "step": 2241
    },
    {
      "epoch": 34.49230769230769,
      "grad_norm": 2.96875,
      "learning_rate": 1.6848970202007836e-05,
      "loss": 0.0227,
      "step": 2242
    },
    {
      "epoch": 34.50769230769231,
      "grad_norm": 11.125,
      "learning_rate": 1.684078769285774e-05,
      "loss": 0.0676,
      "step": 2243
    },
    {
      "epoch": 34.52307692307692,
      "grad_norm": 0.23046875,
      "learning_rate": 1.6832603068060453e-05,
      "loss": 0.0051,
      "step": 2244
    },
    {
      "epoch": 34.53846153846154,
      "grad_norm": 0.62890625,
      "learning_rate": 1.6824416331605062e-05,
      "loss": 0.0127,
      "step": 2245
    },
    {
      "epoch": 34.55384615384615,
      "grad_norm": 0.8828125,
      "learning_rate": 1.6816227487481685e-05,
      "loss": 0.024,
      "step": 2246
    },
    {
      "epoch": 34.56923076923077,
      "grad_norm": 4.84375,
      "learning_rate": 1.6808036539681463e-05,
      "loss": 0.0253,
      "step": 2247
    },
    {
      "epoch": 34.58461538461538,
      "grad_norm": 1.2578125,
      "learning_rate": 1.6799843492196567e-05,
      "loss": 0.0088,
      "step": 2248
    },
    {
      "epoch": 34.6,
      "grad_norm": 2.203125,
      "learning_rate": 1.6791648349020192e-05,
      "loss": 0.0264,
      "step": 2249
    },
    {
      "epoch": 34.61538461538461,
      "grad_norm": 1.4609375,
      "learning_rate": 1.6783451114146545e-05,
      "loss": 0.0133,
      "step": 2250
    },
    {
      "epoch": 34.63076923076923,
      "grad_norm": 4.375,
      "learning_rate": 1.677525179157086e-05,
      "loss": 0.051,
      "step": 2251
    },
    {
      "epoch": 34.646153846153844,
      "grad_norm": 0.314453125,
      "learning_rate": 1.6767050385289386e-05,
      "loss": 0.0044,
      "step": 2252
    },
    {
      "epoch": 34.66153846153846,
      "grad_norm": 1.1171875,
      "learning_rate": 1.67588468992994e-05,
      "loss": 0.0071,
      "step": 2253
    },
    {
      "epoch": 34.676923076923075,
      "grad_norm": 0.9921875,
      "learning_rate": 1.675064133759916e-05,
      "loss": 0.0205,
      "step": 2254
    },
    {
      "epoch": 34.69230769230769,
      "grad_norm": 0.82421875,
      "learning_rate": 1.6742433704187977e-05,
      "loss": 0.0349,
      "step": 2255
    },
    {
      "epoch": 34.707692307692305,
      "grad_norm": 5.375,
      "learning_rate": 1.6734224003066146e-05,
      "loss": 0.0337,
      "step": 2256
    },
    {
      "epoch": 34.723076923076924,
      "grad_norm": 1.8984375,
      "learning_rate": 1.6726012238234967e-05,
      "loss": 0.0224,
      "step": 2257
    },
    {
      "epoch": 34.738461538461536,
      "grad_norm": 2.9375,
      "learning_rate": 1.6717798413696765e-05,
      "loss": 0.0297,
      "step": 2258
    },
    {
      "epoch": 34.753846153846155,
      "grad_norm": 5.03125,
      "learning_rate": 1.6709582533454853e-05,
      "loss": 0.0242,
      "step": 2259
    },
    {
      "epoch": 34.76923076923077,
      "grad_norm": 0.75,
      "learning_rate": 1.6701364601513557e-05,
      "loss": 0.0097,
      "step": 2260
    },
    {
      "epoch": 34.784615384615385,
      "grad_norm": 1.25,
      "learning_rate": 1.669314462187819e-05,
      "loss": 0.0222,
      "step": 2261
    },
    {
      "epoch": 34.8,
      "grad_norm": 0.83203125,
      "learning_rate": 1.6684922598555074e-05,
      "loss": 0.02,
      "step": 2262
    },
    {
      "epoch": 34.815384615384616,
      "grad_norm": 1.1171875,
      "learning_rate": 1.6676698535551523e-05,
      "loss": 0.0731,
      "step": 2263
    },
    {
      "epoch": 34.83076923076923,
      "grad_norm": 1.9375,
      "learning_rate": 1.666847243687584e-05,
      "loss": 0.0148,
      "step": 2264
    },
    {
      "epoch": 34.84615384615385,
      "grad_norm": 1.8515625,
      "learning_rate": 1.6660244306537335e-05,
      "loss": 0.0199,
      "step": 2265
    },
    {
      "epoch": 34.86153846153846,
      "grad_norm": 5.0625,
      "learning_rate": 1.665201414854629e-05,
      "loss": 0.0434,
      "step": 2266
    },
    {
      "epoch": 34.87692307692308,
      "grad_norm": 3.515625,
      "learning_rate": 1.6643781966913997e-05,
      "loss": 0.0309,
      "step": 2267
    },
    {
      "epoch": 34.89230769230769,
      "grad_norm": 1.46875,
      "learning_rate": 1.6635547765652704e-05,
      "loss": 0.0215,
      "step": 2268
    },
    {
      "epoch": 34.90769230769231,
      "grad_norm": 1.2421875,
      "learning_rate": 1.662731154877567e-05,
      "loss": 0.0071,
      "step": 2269
    },
    {
      "epoch": 34.92307692307692,
      "grad_norm": 1.078125,
      "learning_rate": 1.6619073320297128e-05,
      "loss": 0.0069,
      "step": 2270
    },
    {
      "epoch": 34.93846153846154,
      "grad_norm": 0.6875,
      "learning_rate": 1.661083308423229e-05,
      "loss": 0.0092,
      "step": 2271
    },
    {
      "epoch": 34.95384615384615,
      "grad_norm": 1.328125,
      "learning_rate": 1.660259084459735e-05,
      "loss": 0.017,
      "step": 2272
    },
    {
      "epoch": 34.96923076923077,
      "grad_norm": 0.6953125,
      "learning_rate": 1.659434660540946e-05,
      "loss": 0.0067,
      "step": 2273
    },
    {
      "epoch": 34.98461538461538,
      "grad_norm": 1.546875,
      "learning_rate": 1.6586100370686782e-05,
      "loss": 0.0239,
      "step": 2274
    },
    {
      "epoch": 35.0,
      "grad_norm": 1.140625,
      "learning_rate": 1.657785214444842e-05,
      "loss": 0.0108,
      "step": 2275
    },
    {
      "epoch": 35.01538461538462,
      "grad_norm": 0.6875,
      "learning_rate": 1.656960193071446e-05,
      "loss": 0.0237,
      "step": 2276
    },
    {
      "epoch": 35.03076923076923,
      "grad_norm": 1.84375,
      "learning_rate": 1.6561349733505956e-05,
      "loss": 0.0142,
      "step": 2277
    },
    {
      "epoch": 35.04615384615385,
      "grad_norm": 3.734375,
      "learning_rate": 1.655309555684493e-05,
      "loss": 0.0371,
      "step": 2278
    },
    {
      "epoch": 35.06153846153846,
      "grad_norm": 3.265625,
      "learning_rate": 1.654483940475437e-05,
      "loss": 0.0136,
      "step": 2279
    },
    {
      "epoch": 35.07692307692308,
      "grad_norm": 1.34375,
      "learning_rate": 1.653658128125822e-05,
      "loss": 0.0151,
      "step": 2280
    },
    {
      "epoch": 35.09230769230769,
      "grad_norm": 0.365234375,
      "learning_rate": 1.652832119038139e-05,
      "loss": 0.0405,
      "step": 2281
    },
    {
      "epoch": 35.10769230769231,
      "grad_norm": 3.125,
      "learning_rate": 1.6520059136149747e-05,
      "loss": 0.027,
      "step": 2282
    },
    {
      "epoch": 35.12307692307692,
      "grad_norm": 6.125,
      "learning_rate": 1.651179512259012e-05,
      "loss": 0.0305,
      "step": 2283
    },
    {
      "epoch": 35.13846153846154,
      "grad_norm": 2.34375,
      "learning_rate": 1.6503529153730285e-05,
      "loss": 0.0331,
      "step": 2284
    },
    {
      "epoch": 35.15384615384615,
      "grad_norm": 2.328125,
      "learning_rate": 1.6495261233598976e-05,
      "loss": 0.039,
      "step": 2285
    },
    {
      "epoch": 35.16923076923077,
      "grad_norm": 4.75,
      "learning_rate": 1.6486991366225878e-05,
      "loss": 0.0321,
      "step": 2286
    },
    {
      "epoch": 35.184615384615384,
      "grad_norm": 1.140625,
      "learning_rate": 1.647871955564162e-05,
      "loss": 0.0217,
      "step": 2287
    },
    {
      "epoch": 35.2,
      "grad_norm": 3.53125,
      "learning_rate": 1.6470445805877782e-05,
      "loss": 0.023,
      "step": 2288
    },
    {
      "epoch": 35.215384615384615,
      "grad_norm": 4.375,
      "learning_rate": 1.64621701209669e-05,
      "loss": 0.0421,
      "step": 2289
    },
    {
      "epoch": 35.23076923076923,
      "grad_norm": 5.6875,
      "learning_rate": 1.6453892504942425e-05,
      "loss": 0.0475,
      "step": 2290
    },
    {
      "epoch": 35.246153846153845,
      "grad_norm": 0.57421875,
      "learning_rate": 1.6445612961838783e-05,
      "loss": 0.029,
      "step": 2291
    },
    {
      "epoch": 35.261538461538464,
      "grad_norm": 2.3125,
      "learning_rate": 1.643733149569131e-05,
      "loss": 0.0225,
      "step": 2292
    },
    {
      "epoch": 35.276923076923076,
      "grad_norm": 2.0,
      "learning_rate": 1.64290481105363e-05,
      "loss": 0.0079,
      "step": 2293
    },
    {
      "epoch": 35.292307692307695,
      "grad_norm": 1.75,
      "learning_rate": 1.6420762810410973e-05,
      "loss": 0.0146,
      "step": 2294
    },
    {
      "epoch": 35.30769230769231,
      "grad_norm": 0.7890625,
      "learning_rate": 1.6412475599353477e-05,
      "loss": 0.0122,
      "step": 2295
    },
    {
      "epoch": 35.323076923076925,
      "grad_norm": 1.1953125,
      "learning_rate": 1.6404186481402908e-05,
      "loss": 0.0205,
      "step": 2296
    },
    {
      "epoch": 35.33846153846154,
      "grad_norm": 1.6015625,
      "learning_rate": 1.6395895460599274e-05,
      "loss": 0.0587,
      "step": 2297
    },
    {
      "epoch": 35.353846153846156,
      "grad_norm": 0.4453125,
      "learning_rate": 1.6387602540983525e-05,
      "loss": 0.0269,
      "step": 2298
    },
    {
      "epoch": 35.36923076923077,
      "grad_norm": 0.6015625,
      "learning_rate": 1.6379307726597517e-05,
      "loss": 0.0086,
      "step": 2299
    },
    {
      "epoch": 35.38461538461539,
      "grad_norm": 0.73046875,
      "learning_rate": 1.6371011021484052e-05,
      "loss": 0.0068,
      "step": 2300
    },
    {
      "epoch": 35.4,
      "grad_norm": 2.28125,
      "learning_rate": 1.6362712429686846e-05,
      "loss": 0.0232,
      "step": 2301
    },
    {
      "epoch": 35.41538461538462,
      "grad_norm": 0.71484375,
      "learning_rate": 1.635441195525052e-05,
      "loss": 0.0048,
      "step": 2302
    },
    {
      "epoch": 35.43076923076923,
      "grad_norm": 2.9375,
      "learning_rate": 1.6346109602220636e-05,
      "loss": 0.042,
      "step": 2303
    },
    {
      "epoch": 35.44615384615385,
      "grad_norm": 0.73828125,
      "learning_rate": 1.633780537464365e-05,
      "loss": 0.0252,
      "step": 2304
    },
    {
      "epoch": 35.46153846153846,
      "grad_norm": 0.703125,
      "learning_rate": 1.6329499276566948e-05,
      "loss": 0.0377,
      "step": 2305
    },
    {
      "epoch": 35.47692307692308,
      "grad_norm": 4.03125,
      "learning_rate": 1.632119131203882e-05,
      "loss": 0.0594,
      "step": 2306
    },
    {
      "epoch": 35.49230769230769,
      "grad_norm": 6.65625,
      "learning_rate": 1.6312881485108473e-05,
      "loss": 0.0783,
      "step": 2307
    },
    {
      "epoch": 35.50769230769231,
      "grad_norm": 1.234375,
      "learning_rate": 1.6304569799826004e-05,
      "loss": 0.0118,
      "step": 2308
    },
    {
      "epoch": 35.52307692307692,
      "grad_norm": 2.6875,
      "learning_rate": 1.6296256260242428e-05,
      "loss": 0.0391,
      "step": 2309
    },
    {
      "epoch": 35.53846153846154,
      "grad_norm": 0.5859375,
      "learning_rate": 1.6287940870409677e-05,
      "loss": 0.0102,
      "step": 2310
    },
    {
      "epoch": 35.55384615384615,
      "grad_norm": 4.4375,
      "learning_rate": 1.6279623634380557e-05,
      "loss": 0.0625,
      "step": 2311
    },
    {
      "epoch": 35.56923076923077,
      "grad_norm": 3.546875,
      "learning_rate": 1.6271304556208795e-05,
      "loss": 0.0166,
      "step": 2312
    },
    {
      "epoch": 35.58461538461538,
      "grad_norm": 2.109375,
      "learning_rate": 1.6262983639949008e-05,
      "loss": 0.0165,
      "step": 2313
    },
    {
      "epoch": 35.6,
      "grad_norm": 4.25,
      "learning_rate": 1.62546608896567e-05,
      "loss": 0.0237,
      "step": 2314
    },
    {
      "epoch": 35.61538461538461,
      "grad_norm": 1.21875,
      "learning_rate": 1.62463363093883e-05,
      "loss": 0.0118,
      "step": 2315
    },
    {
      "epoch": 35.63076923076923,
      "grad_norm": 1.578125,
      "learning_rate": 1.6238009903201085e-05,
      "loss": 0.0474,
      "step": 2316
    },
    {
      "epoch": 35.646153846153844,
      "grad_norm": 0.59375,
      "learning_rate": 1.6229681675153255e-05,
      "loss": 0.0097,
      "step": 2317
    },
    {
      "epoch": 35.66153846153846,
      "grad_norm": 3.5,
      "learning_rate": 1.6221351629303883e-05,
      "loss": 0.0652,
      "step": 2318
    },
    {
      "epoch": 35.676923076923075,
      "grad_norm": 1.0390625,
      "learning_rate": 1.6213019769712938e-05,
      "loss": 0.0106,
      "step": 2319
    },
    {
      "epoch": 35.69230769230769,
      "grad_norm": 2.84375,
      "learning_rate": 1.6204686100441264e-05,
      "loss": 0.0323,
      "step": 2320
    },
    {
      "epoch": 35.707692307692305,
      "grad_norm": 0.447265625,
      "learning_rate": 1.6196350625550585e-05,
      "loss": 0.009,
      "step": 2321
    },
    {
      "epoch": 35.723076923076924,
      "grad_norm": 3.59375,
      "learning_rate": 1.618801334910352e-05,
      "loss": 0.033,
      "step": 2322
    },
    {
      "epoch": 35.738461538461536,
      "grad_norm": 2.25,
      "learning_rate": 1.6179674275163545e-05,
      "loss": 0.0351,
      "step": 2323
    },
    {
      "epoch": 35.753846153846155,
      "grad_norm": 1.2109375,
      "learning_rate": 1.6171333407795036e-05,
      "loss": 0.0308,
      "step": 2324
    },
    {
      "epoch": 35.76923076923077,
      "grad_norm": 3.109375,
      "learning_rate": 1.6162990751063217e-05,
      "loss": 0.0123,
      "step": 2325
    },
    {
      "epoch": 35.784615384615385,
      "grad_norm": 1.3828125,
      "learning_rate": 1.615464630903421e-05,
      "loss": 0.0146,
      "step": 2326
    },
    {
      "epoch": 35.8,
      "grad_norm": 2.4375,
      "learning_rate": 1.6146300085774986e-05,
      "loss": 0.0157,
      "step": 2327
    },
    {
      "epoch": 35.815384615384616,
      "grad_norm": 1.9375,
      "learning_rate": 1.6137952085353398e-05,
      "loss": 0.0342,
      "step": 2328
    },
    {
      "epoch": 35.83076923076923,
      "grad_norm": 1.34375,
      "learning_rate": 1.612960231183816e-05,
      "loss": 0.0188,
      "step": 2329
    },
    {
      "epoch": 35.84615384615385,
      "grad_norm": 2.3125,
      "learning_rate": 1.6121250769298844e-05,
      "loss": 0.0147,
      "step": 2330
    },
    {
      "epoch": 35.86153846153846,
      "grad_norm": 5.6875,
      "learning_rate": 1.6112897461805895e-05,
      "loss": 0.0325,
      "step": 2331
    },
    {
      "epoch": 35.87692307692308,
      "grad_norm": 2.59375,
      "learning_rate": 1.6104542393430615e-05,
      "loss": 0.018,
      "step": 2332
    },
    {
      "epoch": 35.89230769230769,
      "grad_norm": 0.546875,
      "learning_rate": 1.609618556824516e-05,
      "loss": 0.0175,
      "step": 2333
    },
    {
      "epoch": 35.90769230769231,
      "grad_norm": 0.56640625,
      "learning_rate": 1.6087826990322546e-05,
      "loss": 0.0094,
      "step": 2334
    },
    {
      "epoch": 35.92307692307692,
      "grad_norm": 5.59375,
      "learning_rate": 1.607946666373664e-05,
      "loss": 0.056,
      "step": 2335
    },
    {
      "epoch": 35.93846153846154,
      "grad_norm": 2.5625,
      "learning_rate": 1.607110459256217e-05,
      "loss": 0.0091,
      "step": 2336
    },
    {
      "epoch": 35.95384615384615,
      "grad_norm": 5.125,
      "learning_rate": 1.60627407808747e-05,
      "loss": 0.0692,
      "step": 2337
    },
    {
      "epoch": 35.96923076923077,
      "grad_norm": 5.46875,
      "learning_rate": 1.605437523275066e-05,
      "loss": 0.0444,
      "step": 2338
    },
    {
      "epoch": 35.98461538461538,
      "grad_norm": 0.2275390625,
      "learning_rate": 1.6046007952267305e-05,
      "loss": 0.0086,
      "step": 2339
    },
    {
      "epoch": 36.0,
      "grad_norm": 1.7265625,
      "learning_rate": 1.6037638943502754e-05,
      "loss": 0.0298,
      "step": 2340
    },
    {
      "epoch": 36.01538461538462,
      "grad_norm": 1.8984375,
      "learning_rate": 1.602926821053596e-05,
      "loss": 0.0186,
      "step": 2341
    },
    {
      "epoch": 36.03076923076923,
      "grad_norm": 3.65625,
      "learning_rate": 1.6020895757446708e-05,
      "loss": 0.0464,
      "step": 2342
    },
    {
      "epoch": 36.04615384615385,
      "grad_norm": 1.1875,
      "learning_rate": 1.601252158831564e-05,
      "loss": 0.0074,
      "step": 2343
    },
    {
      "epoch": 36.06153846153846,
      "grad_norm": 2.046875,
      "learning_rate": 1.600414570722422e-05,
      "loss": 0.0108,
      "step": 2344
    },
    {
      "epoch": 36.07692307692308,
      "grad_norm": 0.427734375,
      "learning_rate": 1.599576811825475e-05,
      "loss": 0.0044,
      "step": 2345
    },
    {
      "epoch": 36.09230769230769,
      "grad_norm": 0.78515625,
      "learning_rate": 1.5987388825490368e-05,
      "loss": 0.0761,
      "step": 2346
    },
    {
      "epoch": 36.10769230769231,
      "grad_norm": 3.28125,
      "learning_rate": 1.5979007833015036e-05,
      "loss": 0.0413,
      "step": 2347
    },
    {
      "epoch": 36.12307692307692,
      "grad_norm": 0.89453125,
      "learning_rate": 1.5970625144913552e-05,
      "loss": 0.0113,
      "step": 2348
    },
    {
      "epoch": 36.13846153846154,
      "grad_norm": 2.859375,
      "learning_rate": 1.596224076527153e-05,
      "loss": 0.0362,
      "step": 2349
    },
    {
      "epoch": 36.15384615384615,
      "grad_norm": 3.5625,
      "learning_rate": 1.5953854698175422e-05,
      "loss": 0.0179,
      "step": 2350
    },
    {
      "epoch": 36.16923076923077,
      "grad_norm": 1.921875,
      "learning_rate": 1.5945466947712488e-05,
      "loss": 0.0136,
      "step": 2351
    },
    {
      "epoch": 36.184615384615384,
      "grad_norm": 3.609375,
      "learning_rate": 1.593707751797083e-05,
      "loss": 0.0291,
      "step": 2352
    },
    {
      "epoch": 36.2,
      "grad_norm": 5.90625,
      "learning_rate": 1.5928686413039334e-05,
      "loss": 0.0382,
      "step": 2353
    },
    {
      "epoch": 36.215384615384615,
      "grad_norm": 1.046875,
      "learning_rate": 1.592029363700773e-05,
      "loss": 0.0142,
      "step": 2354
    },
    {
      "epoch": 36.23076923076923,
      "grad_norm": 0.5859375,
      "learning_rate": 1.5911899193966567e-05,
      "loss": 0.0111,
      "step": 2355
    },
    {
      "epoch": 36.246153846153845,
      "grad_norm": 0.5078125,
      "learning_rate": 1.590350308800718e-05,
      "loss": 0.0066,
      "step": 2356
    },
    {
      "epoch": 36.261538461538464,
      "grad_norm": 4.09375,
      "learning_rate": 1.5895105323221735e-05,
      "loss": 0.0398,
      "step": 2357
    },
    {
      "epoch": 36.276923076923076,
      "grad_norm": 0.3671875,
      "learning_rate": 1.58867059037032e-05,
      "loss": 0.0044,
      "step": 2358
    },
    {
      "epoch": 36.292307692307695,
      "grad_norm": 2.5625,
      "learning_rate": 1.5878304833545348e-05,
      "loss": 0.0301,
      "step": 2359
    },
    {
      "epoch": 36.30769230769231,
      "grad_norm": 1.6015625,
      "learning_rate": 1.5869902116842765e-05,
      "loss": 0.0262,
      "step": 2360
    },
    {
      "epoch": 36.323076923076925,
      "grad_norm": 1.34375,
      "learning_rate": 1.586149775769082e-05,
      "loss": 0.009,
      "step": 2361
    },
    {
      "epoch": 36.33846153846154,
      "grad_norm": 2.328125,
      "learning_rate": 1.585309176018571e-05,
      "loss": 0.0217,
      "step": 2362
    },
    {
      "epoch": 36.353846153846156,
      "grad_norm": 1.109375,
      "learning_rate": 1.584468412842441e-05,
      "loss": 0.0075,
      "step": 2363
    },
    {
      "epoch": 36.36923076923077,
      "grad_norm": 0.78515625,
      "learning_rate": 1.5836274866504698e-05,
      "loss": 0.0144,
      "step": 2364
    },
    {
      "epoch": 36.38461538461539,
      "grad_norm": 1.421875,
      "learning_rate": 1.5827863978525143e-05,
      "loss": 0.0131,
      "step": 2365
    },
    {
      "epoch": 36.4,
      "grad_norm": 0.60546875,
      "learning_rate": 1.581945146858511e-05,
      "loss": 0.0231,
      "step": 2366
    },
    {
      "epoch": 36.41538461538462,
      "grad_norm": 2.546875,
      "learning_rate": 1.581103734078476e-05,
      "loss": 0.0491,
      "step": 2367
    },
    {
      "epoch": 36.43076923076923,
      "grad_norm": 0.75390625,
      "learning_rate": 1.5802621599225033e-05,
      "loss": 0.0235,
      "step": 2368
    },
    {
      "epoch": 36.44615384615385,
      "grad_norm": 0.3359375,
      "learning_rate": 1.579420424800766e-05,
      "loss": 0.0569,
      "step": 2369
    },
    {
      "epoch": 36.46153846153846,
      "grad_norm": 2.1875,
      "learning_rate": 1.578578529123515e-05,
      "loss": 0.014,
      "step": 2370
    },
    {
      "epoch": 36.47692307692308,
      "grad_norm": 2.078125,
      "learning_rate": 1.577736473301081e-05,
      "loss": 0.0372,
      "step": 2371
    },
    {
      "epoch": 36.49230769230769,
      "grad_norm": 3.015625,
      "learning_rate": 1.576894257743871e-05,
      "loss": 0.0431,
      "step": 2372
    },
    {
      "epoch": 36.50769230769231,
      "grad_norm": 3.953125,
      "learning_rate": 1.5760518828623715e-05,
      "loss": 0.0189,
      "step": 2373
    },
    {
      "epoch": 36.52307692307692,
      "grad_norm": 2.015625,
      "learning_rate": 1.5752093490671447e-05,
      "loss": 0.0228,
      "step": 2374
    },
    {
      "epoch": 36.53846153846154,
      "grad_norm": 2.453125,
      "learning_rate": 1.574366656768832e-05,
      "loss": 0.0291,
      "step": 2375
    },
    {
      "epoch": 36.55384615384615,
      "grad_norm": 4.875,
      "learning_rate": 1.573523806378151e-05,
      "loss": 0.0216,
      "step": 2376
    },
    {
      "epoch": 36.56923076923077,
      "grad_norm": 5.71875,
      "learning_rate": 1.5726807983058974e-05,
      "loss": 0.0477,
      "step": 2377
    },
    {
      "epoch": 36.58461538461538,
      "grad_norm": 4.0,
      "learning_rate": 1.5718376329629424e-05,
      "loss": 0.0166,
      "step": 2378
    },
    {
      "epoch": 36.6,
      "grad_norm": 1.59375,
      "learning_rate": 1.5709943107602347e-05,
      "loss": 0.0385,
      "step": 2379
    },
    {
      "epoch": 36.61538461538461,
      "grad_norm": 2.109375,
      "learning_rate": 1.5701508321087995e-05,
      "loss": 0.0294,
      "step": 2380
    },
    {
      "epoch": 36.63076923076923,
      "grad_norm": 1.0546875,
      "learning_rate": 1.569307197419738e-05,
      "loss": 0.0096,
      "step": 2381
    },
    {
      "epoch": 36.646153846153844,
      "grad_norm": 3.09375,
      "learning_rate": 1.5684634071042278e-05,
      "loss": 0.0089,
      "step": 2382
    },
    {
      "epoch": 36.66153846153846,
      "grad_norm": 2.015625,
      "learning_rate": 1.5676194615735217e-05,
      "loss": 0.0187,
      "step": 2383
    },
    {
      "epoch": 36.676923076923075,
      "grad_norm": 1.3671875,
      "learning_rate": 1.566775361238948e-05,
      "loss": 0.0093,
      "step": 2384
    },
    {
      "epoch": 36.69230769230769,
      "grad_norm": 10.1875,
      "learning_rate": 1.5659311065119118e-05,
      "loss": 0.0781,
      "step": 2385
    },
    {
      "epoch": 36.707692307692305,
      "grad_norm": 2.28125,
      "learning_rate": 1.5650866978038925e-05,
      "loss": 0.0178,
      "step": 2386
    },
    {
      "epoch": 36.723076923076924,
      "grad_norm": 2.75,
      "learning_rate": 1.5642421355264438e-05,
      "loss": 0.0147,
      "step": 2387
    },
    {
      "epoch": 36.738461538461536,
      "grad_norm": 1.0,
      "learning_rate": 1.563397420091196e-05,
      "loss": 0.0157,
      "step": 2388
    },
    {
      "epoch": 36.753846153846155,
      "grad_norm": 5.09375,
      "learning_rate": 1.5625525519098527e-05,
      "loss": 0.0192,
      "step": 2389
    },
    {
      "epoch": 36.76923076923077,
      "grad_norm": 2.984375,
      "learning_rate": 1.5617075313941925e-05,
      "loss": 0.0348,
      "step": 2390
    },
    {
      "epoch": 36.784615384615385,
      "grad_norm": 2.84375,
      "learning_rate": 1.560862358956069e-05,
      "loss": 0.024,
      "step": 2391
    },
    {
      "epoch": 36.8,
      "grad_norm": 1.59375,
      "learning_rate": 1.560017035007407e-05,
      "loss": 0.0125,
      "step": 2392
    },
    {
      "epoch": 36.815384615384616,
      "grad_norm": 2.5,
      "learning_rate": 1.559171559960209e-05,
      "loss": 0.0336,
      "step": 2393
    },
    {
      "epoch": 36.83076923076923,
      "grad_norm": 1.15625,
      "learning_rate": 1.5583259342265483e-05,
      "loss": 0.0112,
      "step": 2394
    },
    {
      "epoch": 36.84615384615385,
      "grad_norm": 0.6015625,
      "learning_rate": 1.5574801582185726e-05,
      "loss": 0.0237,
      "step": 2395
    },
    {
      "epoch": 36.86153846153846,
      "grad_norm": 1.3046875,
      "learning_rate": 1.5566342323485032e-05,
      "loss": 0.0101,
      "step": 2396
    },
    {
      "epoch": 36.87692307692308,
      "grad_norm": 3.546875,
      "learning_rate": 1.555788157028634e-05,
      "loss": 0.0243,
      "step": 2397
    },
    {
      "epoch": 36.89230769230769,
      "grad_norm": 3.40625,
      "learning_rate": 1.5549419326713323e-05,
      "loss": 0.0291,
      "step": 2398
    },
    {
      "epoch": 36.90769230769231,
      "grad_norm": 0.9609375,
      "learning_rate": 1.5540955596890365e-05,
      "loss": 0.008,
      "step": 2399
    },
    {
      "epoch": 36.92307692307692,
      "grad_norm": 2.28125,
      "learning_rate": 1.5532490384942593e-05,
      "loss": 0.0151,
      "step": 2400
    },
    {
      "epoch": 36.93846153846154,
      "grad_norm": 1.296875,
      "learning_rate": 1.5524023694995844e-05,
      "loss": 0.0144,
      "step": 2401
    },
    {
      "epoch": 36.95384615384615,
      "grad_norm": 1.4140625,
      "learning_rate": 1.5515555531176687e-05,
      "loss": 0.023,
      "step": 2402
    },
    {
      "epoch": 36.96923076923077,
      "grad_norm": 2.09375,
      "learning_rate": 1.5507085897612396e-05,
      "loss": 0.0093,
      "step": 2403
    },
    {
      "epoch": 36.98461538461538,
      "grad_norm": 2.28125,
      "learning_rate": 1.5498614798430972e-05,
      "loss": 0.0103,
      "step": 2404
    },
    {
      "epoch": 37.0,
      "grad_norm": 0.314453125,
      "learning_rate": 1.549014223776112e-05,
      "loss": 0.0026,
      "step": 2405
    },
    {
      "epoch": 37.01538461538462,
      "grad_norm": 0.38671875,
      "learning_rate": 1.5481668219732263e-05,
      "loss": 0.0043,
      "step": 2406
    },
    {
      "epoch": 37.03076923076923,
      "grad_norm": 2.875,
      "learning_rate": 1.547319274847454e-05,
      "loss": 0.0387,
      "step": 2407
    },
    {
      "epoch": 37.04615384615385,
      "grad_norm": 4.71875,
      "learning_rate": 1.546471582811879e-05,
      "loss": 0.0456,
      "step": 2408
    },
    {
      "epoch": 37.06153846153846,
      "grad_norm": 1.734375,
      "learning_rate": 1.545623746279656e-05,
      "loss": 0.0134,
      "step": 2409
    },
    {
      "epoch": 37.07692307692308,
      "grad_norm": 1.765625,
      "learning_rate": 1.5447757656640104e-05,
      "loss": 0.0177,
      "step": 2410
    },
    {
      "epoch": 37.09230769230769,
      "grad_norm": 1.734375,
      "learning_rate": 1.5439276413782375e-05,
      "loss": 0.0184,
      "step": 2411
    },
    {
      "epoch": 37.10769230769231,
      "grad_norm": 4.34375,
      "learning_rate": 1.543079373835703e-05,
      "loss": 0.0455,
      "step": 2412
    },
    {
      "epoch": 37.12307692307692,
      "grad_norm": 14.25,
      "learning_rate": 1.5422309634498423e-05,
      "loss": 0.0274,
      "step": 2413
    },
    {
      "epoch": 37.13846153846154,
      "grad_norm": 4.09375,
      "learning_rate": 1.54138241063416e-05,
      "loss": 0.0284,
      "step": 2414
    },
    {
      "epoch": 37.15384615384615,
      "grad_norm": 1.25,
      "learning_rate": 1.54053371580223e-05,
      "loss": 0.0078,
      "step": 2415
    },
    {
      "epoch": 37.16923076923077,
      "grad_norm": 0.609375,
      "learning_rate": 1.5396848793676966e-05,
      "loss": 0.0112,
      "step": 2416
    },
    {
      "epoch": 37.184615384615384,
      "grad_norm": 0.7578125,
      "learning_rate": 1.5388359017442727e-05,
      "loss": 0.0096,
      "step": 2417
    },
    {
      "epoch": 37.2,
      "grad_norm": 3.046875,
      "learning_rate": 1.5379867833457386e-05,
      "loss": 0.0365,
      "step": 2418
    },
    {
      "epoch": 37.215384615384615,
      "grad_norm": 2.875,
      "learning_rate": 1.5371375245859446e-05,
      "loss": 0.0139,
      "step": 2419
    },
    {
      "epoch": 37.23076923076923,
      "grad_norm": 1.578125,
      "learning_rate": 1.5362881258788098e-05,
      "loss": 0.0142,
      "step": 2420
    },
    {
      "epoch": 37.246153846153845,
      "grad_norm": 2.984375,
      "learning_rate": 1.5354385876383197e-05,
      "loss": 0.0371,
      "step": 2421
    },
    {
      "epoch": 37.261538461538464,
      "grad_norm": 2.78125,
      "learning_rate": 1.53458891027853e-05,
      "loss": 0.0267,
      "step": 2422
    },
    {
      "epoch": 37.276923076923076,
      "grad_norm": 5.125,
      "learning_rate": 1.5337390942135627e-05,
      "loss": 0.0461,
      "step": 2423
    },
    {
      "epoch": 37.292307692307695,
      "grad_norm": 0.9296875,
      "learning_rate": 1.5328891398576077e-05,
      "loss": 0.0865,
      "step": 2424
    },
    {
      "epoch": 37.30769230769231,
      "grad_norm": 0.64453125,
      "learning_rate": 1.532039047624922e-05,
      "loss": 0.0135,
      "step": 2425
    },
    {
      "epoch": 37.323076923076925,
      "grad_norm": 3.25,
      "learning_rate": 1.5311888179298313e-05,
      "loss": 0.0253,
      "step": 2426
    },
    {
      "epoch": 37.33846153846154,
      "grad_norm": 2.0,
      "learning_rate": 1.5303384511867268e-05,
      "loss": 0.0121,
      "step": 2427
    },
    {
      "epoch": 37.353846153846156,
      "grad_norm": 4.0,
      "learning_rate": 1.5294879478100663e-05,
      "loss": 0.0562,
      "step": 2428
    },
    {
      "epoch": 37.36923076923077,
      "grad_norm": 7.09375,
      "learning_rate": 1.5286373082143757e-05,
      "loss": 0.0373,
      "step": 2429
    },
    {
      "epoch": 37.38461538461539,
      "grad_norm": 8.0,
      "learning_rate": 1.5277865328142457e-05,
      "loss": 0.0304,
      "step": 2430
    },
    {
      "epoch": 37.4,
      "grad_norm": 3.53125,
      "learning_rate": 1.5269356220243345e-05,
      "loss": 0.0304,
      "step": 2431
    },
    {
      "epoch": 37.41538461538462,
      "grad_norm": 1.6328125,
      "learning_rate": 1.5260845762593647e-05,
      "loss": 0.011,
      "step": 2432
    },
    {
      "epoch": 37.43076923076923,
      "grad_norm": 1.5703125,
      "learning_rate": 1.525233395934127e-05,
      "loss": 0.0258,
      "step": 2433
    },
    {
      "epoch": 37.44615384615385,
      "grad_norm": 1.5625,
      "learning_rate": 1.5243820814634758e-05,
      "loss": 0.0386,
      "step": 2434
    },
    {
      "epoch": 37.46153846153846,
      "grad_norm": 0.80078125,
      "learning_rate": 1.5235306332623309e-05,
      "loss": 0.0067,
      "step": 2435
    },
    {
      "epoch": 37.47692307692308,
      "grad_norm": 1.234375,
      "learning_rate": 1.5226790517456785e-05,
      "loss": 0.0223,
      "step": 2436
    },
    {
      "epoch": 37.49230769230769,
      "grad_norm": 3.34375,
      "learning_rate": 1.5218273373285686e-05,
      "loss": 0.0179,
      "step": 2437
    },
    {
      "epoch": 37.50769230769231,
      "grad_norm": 2.890625,
      "learning_rate": 1.5209754904261167e-05,
      "loss": 0.0311,
      "step": 2438
    },
    {
      "epoch": 37.52307692307692,
      "grad_norm": 2.359375,
      "learning_rate": 1.5201235114535028e-05,
      "loss": 0.0318,
      "step": 2439
    },
    {
      "epoch": 37.53846153846154,
      "grad_norm": 0.609375,
      "learning_rate": 1.5192714008259706e-05,
      "loss": 0.0176,
      "step": 2440
    },
    {
      "epoch": 37.55384615384615,
      "grad_norm": 5.15625,
      "learning_rate": 1.5184191589588292e-05,
      "loss": 0.0382,
      "step": 2441
    },
    {
      "epoch": 37.56923076923077,
      "grad_norm": 0.546875,
      "learning_rate": 1.51756678626745e-05,
      "loss": 0.0045,
      "step": 2442
    },
    {
      "epoch": 37.58461538461538,
      "grad_norm": 2.84375,
      "learning_rate": 1.5167142831672702e-05,
      "loss": 0.0242,
      "step": 2443
    },
    {
      "epoch": 37.6,
      "grad_norm": 3.65625,
      "learning_rate": 1.5158616500737885e-05,
      "loss": 0.0211,
      "step": 2444
    },
    {
      "epoch": 37.61538461538461,
      "grad_norm": 0.9609375,
      "learning_rate": 1.5150088874025686e-05,
      "loss": 0.0241,
      "step": 2445
    },
    {
      "epoch": 37.63076923076923,
      "grad_norm": 2.75,
      "learning_rate": 1.514155995569236e-05,
      "loss": 0.0214,
      "step": 2446
    },
    {
      "epoch": 37.646153846153844,
      "grad_norm": 0.76171875,
      "learning_rate": 1.51330297498948e-05,
      "loss": 0.0117,
      "step": 2447
    },
    {
      "epoch": 37.66153846153846,
      "grad_norm": 1.1640625,
      "learning_rate": 1.5124498260790535e-05,
      "loss": 0.0074,
      "step": 2448
    },
    {
      "epoch": 37.676923076923075,
      "grad_norm": 3.546875,
      "learning_rate": 1.511596549253769e-05,
      "loss": 0.0344,
      "step": 2449
    },
    {
      "epoch": 37.69230769230769,
      "grad_norm": 2.140625,
      "learning_rate": 1.5107431449295047e-05,
      "loss": 0.0312,
      "step": 2450
    },
    {
      "epoch": 37.707692307692305,
      "grad_norm": 2.28125,
      "learning_rate": 1.5098896135221991e-05,
      "loss": 0.0384,
      "step": 2451
    },
    {
      "epoch": 37.723076923076924,
      "grad_norm": 4.15625,
      "learning_rate": 1.5090359554478531e-05,
      "loss": 0.0383,
      "step": 2452
    },
    {
      "epoch": 37.738461538461536,
      "grad_norm": 1.8046875,
      "learning_rate": 1.5081821711225295e-05,
      "loss": 0.0451,
      "step": 2453
    },
    {
      "epoch": 37.753846153846155,
      "grad_norm": 6.40625,
      "learning_rate": 1.5073282609623518e-05,
      "loss": 0.063,
      "step": 2454
    },
    {
      "epoch": 37.76923076923077,
      "grad_norm": 0.265625,
      "learning_rate": 1.5064742253835057e-05,
      "loss": 0.0064,
      "step": 2455
    },
    {
      "epoch": 37.784615384615385,
      "grad_norm": 9.75,
      "learning_rate": 1.5056200648022378e-05,
      "loss": 0.066,
      "step": 2456
    },
    {
      "epoch": 37.8,
      "grad_norm": 1.9765625,
      "learning_rate": 1.5047657796348563e-05,
      "loss": 0.0104,
      "step": 2457
    },
    {
      "epoch": 37.815384615384616,
      "grad_norm": 3.015625,
      "learning_rate": 1.5039113702977278e-05,
      "loss": 0.0199,
      "step": 2458
    },
    {
      "epoch": 37.83076923076923,
      "grad_norm": 1.328125,
      "learning_rate": 1.5030568372072826e-05,
      "loss": 0.0383,
      "step": 2459
    },
    {
      "epoch": 37.84615384615385,
      "grad_norm": 0.625,
      "learning_rate": 1.5022021807800087e-05,
      "loss": 0.0059,
      "step": 2460
    },
    {
      "epoch": 37.86153846153846,
      "grad_norm": 3.6875,
      "learning_rate": 1.501347401432456e-05,
      "loss": 0.0417,
      "step": 2461
    },
    {
      "epoch": 37.87692307692308,
      "grad_norm": 0.88671875,
      "learning_rate": 1.5004924995812331e-05,
      "loss": 0.0124,
      "step": 2462
    },
    {
      "epoch": 37.89230769230769,
      "grad_norm": 1.375,
      "learning_rate": 1.4996374756430087e-05,
      "loss": 0.0045,
      "step": 2463
    },
    {
      "epoch": 37.90769230769231,
      "grad_norm": 1.3359375,
      "learning_rate": 1.4987823300345117e-05,
      "loss": 0.0135,
      "step": 2464
    },
    {
      "epoch": 37.92307692307692,
      "grad_norm": 1.921875,
      "learning_rate": 1.4979270631725292e-05,
      "loss": 0.0368,
      "step": 2465
    },
    {
      "epoch": 37.93846153846154,
      "grad_norm": 1.4453125,
      "learning_rate": 1.4970716754739078e-05,
      "loss": 0.0112,
      "step": 2466
    },
    {
      "epoch": 37.95384615384615,
      "grad_norm": 1.1875,
      "learning_rate": 1.4962161673555537e-05,
      "loss": 0.0139,
      "step": 2467
    },
    {
      "epoch": 37.96923076923077,
      "grad_norm": 0.55078125,
      "learning_rate": 1.49536053923443e-05,
      "loss": 0.0096,
      "step": 2468
    },
    {
      "epoch": 37.98461538461538,
      "grad_norm": 2.53125,
      "learning_rate": 1.4945047915275604e-05,
      "loss": 0.0223,
      "step": 2469
    },
    {
      "epoch": 38.0,
      "grad_norm": 2.515625,
      "learning_rate": 1.4936489246520257e-05,
      "loss": 0.0135,
      "step": 2470
    },
    {
      "epoch": 38.01538461538462,
      "grad_norm": 1.6640625,
      "learning_rate": 1.492792939024965e-05,
      "loss": 0.0237,
      "step": 2471
    },
    {
      "epoch": 38.03076923076923,
      "grad_norm": 3.59375,
      "learning_rate": 1.4919368350635757e-05,
      "loss": 0.0459,
      "step": 2472
    },
    {
      "epoch": 38.04615384615385,
      "grad_norm": 0.54296875,
      "learning_rate": 1.4910806131851113e-05,
      "loss": 0.0104,
      "step": 2473
    },
    {
      "epoch": 38.06153846153846,
      "grad_norm": 1.5390625,
      "learning_rate": 1.4902242738068853e-05,
      "loss": 0.0206,
      "step": 2474
    },
    {
      "epoch": 38.07692307692308,
      "grad_norm": 2.96875,
      "learning_rate": 1.4893678173462664e-05,
      "loss": 0.0273,
      "step": 2475
    },
    {
      "epoch": 38.09230769230769,
      "grad_norm": 1.2265625,
      "learning_rate": 1.4885112442206812e-05,
      "loss": 0.0511,
      "step": 2476
    },
    {
      "epoch": 38.10769230769231,
      "grad_norm": 1.71875,
      "learning_rate": 1.4876545548476128e-05,
      "loss": 0.034,
      "step": 2477
    },
    {
      "epoch": 38.12307692307692,
      "grad_norm": 1.34375,
      "learning_rate": 1.4867977496446017e-05,
      "loss": 0.017,
      "step": 2478
    },
    {
      "epoch": 38.13846153846154,
      "grad_norm": 1.1484375,
      "learning_rate": 1.4859408290292442e-05,
      "loss": 0.0086,
      "step": 2479
    },
    {
      "epoch": 38.15384615384615,
      "grad_norm": 3.875,
      "learning_rate": 1.4850837934191924e-05,
      "loss": 0.02,
      "step": 2480
    },
    {
      "epoch": 38.16923076923077,
      "grad_norm": 2.75,
      "learning_rate": 1.4842266432321561e-05,
      "loss": 0.0116,
      "step": 2481
    },
    {
      "epoch": 38.184615384615384,
      "grad_norm": 3.390625,
      "learning_rate": 1.4833693788858988e-05,
      "loss": 0.0289,
      "step": 2482
    },
    {
      "epoch": 38.2,
      "grad_norm": 2.609375,
      "learning_rate": 1.4825120007982413e-05,
      "loss": 0.036,
      "step": 2483
    },
    {
      "epoch": 38.215384615384615,
      "grad_norm": 2.015625,
      "learning_rate": 1.4816545093870593e-05,
      "loss": 0.0114,
      "step": 2484
    },
    {
      "epoch": 38.23076923076923,
      "grad_norm": 1.1796875,
      "learning_rate": 1.4807969050702838e-05,
      "loss": 0.0093,
      "step": 2485
    },
    {
      "epoch": 38.246153846153845,
      "grad_norm": 0.291015625,
      "learning_rate": 1.4799391882659004e-05,
      "loss": 0.0145,
      "step": 2486
    },
    {
      "epoch": 38.261538461538464,
      "grad_norm": 2.015625,
      "learning_rate": 1.4790813593919497e-05,
      "loss": 0.0133,
      "step": 2487
    },
    {
      "epoch": 38.276923076923076,
      "grad_norm": 0.5625,
      "learning_rate": 1.4782234188665278e-05,
      "loss": 0.0058,
      "step": 2488
    },
    {
      "epoch": 38.292307692307695,
      "grad_norm": 7.28125,
      "learning_rate": 1.4773653671077836e-05,
      "loss": 0.0309,
      "step": 2489
    },
    {
      "epoch": 38.30769230769231,
      "grad_norm": 0.4140625,
      "learning_rate": 1.4765072045339218e-05,
      "loss": 0.0118,
      "step": 2490
    },
    {
      "epoch": 38.323076923076925,
      "grad_norm": 0.734375,
      "learning_rate": 1.4756489315632003e-05,
      "loss": 0.0079,
      "step": 2491
    },
    {
      "epoch": 38.33846153846154,
      "grad_norm": 1.625,
      "learning_rate": 1.4747905486139305e-05,
      "loss": 0.0206,
      "step": 2492
    },
    {
      "epoch": 38.353846153846156,
      "grad_norm": 3.265625,
      "learning_rate": 1.4739320561044781e-05,
      "loss": 0.024,
      "step": 2493
    },
    {
      "epoch": 38.36923076923077,
      "grad_norm": 0.75390625,
      "learning_rate": 1.4730734544532617e-05,
      "loss": 0.0083,
      "step": 2494
    },
    {
      "epoch": 38.38461538461539,
      "grad_norm": 3.046875,
      "learning_rate": 1.472214744078754e-05,
      "loss": 0.0363,
      "step": 2495
    },
    {
      "epoch": 38.4,
      "grad_norm": 1.171875,
      "learning_rate": 1.4713559253994793e-05,
      "loss": 0.0346,
      "step": 2496
    },
    {
      "epoch": 38.41538461538462,
      "grad_norm": 1.6640625,
      "learning_rate": 1.4704969988340156e-05,
      "loss": 0.0195,
      "step": 2497
    },
    {
      "epoch": 38.43076923076923,
      "grad_norm": 0.828125,
      "learning_rate": 1.4696379648009933e-05,
      "loss": 0.0042,
      "step": 2498
    },
    {
      "epoch": 38.44615384615385,
      "grad_norm": 1.0625,
      "learning_rate": 1.4687788237190952e-05,
      "loss": 0.0042,
      "step": 2499
    },
    {
      "epoch": 38.46153846153846,
      "grad_norm": 4.625,
      "learning_rate": 1.4679195760070561e-05,
      "loss": 0.0357,
      "step": 2500
    },
    {
      "epoch": 38.47692307692308,
      "grad_norm": 3.828125,
      "learning_rate": 1.4670602220836633e-05,
      "loss": 0.0159,
      "step": 2501
    },
    {
      "epoch": 38.49230769230769,
      "grad_norm": 0.953125,
      "learning_rate": 1.4662007623677549e-05,
      "loss": 0.0134,
      "step": 2502
    },
    {
      "epoch": 38.50769230769231,
      "grad_norm": 2.53125,
      "learning_rate": 1.4653411972782211e-05,
      "loss": 0.0174,
      "step": 2503
    },
    {
      "epoch": 38.52307692307692,
      "grad_norm": 6.75,
      "learning_rate": 1.4644815272340037e-05,
      "loss": 0.0394,
      "step": 2504
    },
    {
      "epoch": 38.53846153846154,
      "grad_norm": 5.0,
      "learning_rate": 1.4636217526540961e-05,
      "loss": 0.0252,
      "step": 2505
    },
    {
      "epoch": 38.55384615384615,
      "grad_norm": 3.640625,
      "learning_rate": 1.4627618739575408e-05,
      "loss": 0.0203,
      "step": 2506
    },
    {
      "epoch": 38.56923076923077,
      "grad_norm": 1.203125,
      "learning_rate": 1.461901891563433e-05,
      "loss": 0.0142,
      "step": 2507
    },
    {
      "epoch": 38.58461538461538,
      "grad_norm": 1.1953125,
      "learning_rate": 1.4610418058909172e-05,
      "loss": 0.0078,
      "step": 2508
    },
    {
      "epoch": 38.6,
      "grad_norm": 1.0625,
      "learning_rate": 1.4601816173591895e-05,
      "loss": 0.0081,
      "step": 2509
    },
    {
      "epoch": 38.61538461538461,
      "grad_norm": 5.03125,
      "learning_rate": 1.4593213263874944e-05,
      "loss": 0.0199,
      "step": 2510
    },
    {
      "epoch": 38.63076923076923,
      "grad_norm": 0.6953125,
      "learning_rate": 1.458460933395128e-05,
      "loss": 0.0054,
      "step": 2511
    },
    {
      "epoch": 38.646153846153844,
      "grad_norm": 0.1748046875,
      "learning_rate": 1.457600438801435e-05,
      "loss": 0.0102,
      "step": 2512
    },
    {
      "epoch": 38.66153846153846,
      "grad_norm": 2.734375,
      "learning_rate": 1.4567398430258097e-05,
      "loss": 0.0106,
      "step": 2513
    },
    {
      "epoch": 38.676923076923075,
      "grad_norm": 2.90625,
      "learning_rate": 1.4558791464876966e-05,
      "loss": 0.0216,
      "step": 2514
    },
    {
      "epoch": 38.69230769230769,
      "grad_norm": 5.15625,
      "learning_rate": 1.455018349606589e-05,
      "loss": 0.0453,
      "step": 2515
    },
    {
      "epoch": 38.707692307692305,
      "grad_norm": 12.5,
      "learning_rate": 1.454157452802028e-05,
      "loss": 0.0449,
      "step": 2516
    },
    {
      "epoch": 38.723076923076924,
      "grad_norm": 0.498046875,
      "learning_rate": 1.4532964564936047e-05,
      "loss": 0.0053,
      "step": 2517
    },
    {
      "epoch": 38.738461538461536,
      "grad_norm": 3.25,
      "learning_rate": 1.4524353611009577e-05,
      "loss": 0.0165,
      "step": 2518
    },
    {
      "epoch": 38.753846153846155,
      "grad_norm": 1.6484375,
      "learning_rate": 1.451574167043776e-05,
      "loss": 0.0128,
      "step": 2519
    },
    {
      "epoch": 38.76923076923077,
      "grad_norm": 1.7890625,
      "learning_rate": 1.4507128747417931e-05,
      "loss": 0.0262,
      "step": 2520
    },
    {
      "epoch": 38.784615384615385,
      "grad_norm": 4.1875,
      "learning_rate": 1.4498514846147937e-05,
      "loss": 0.0271,
      "step": 2521
    },
    {
      "epoch": 38.8,
      "grad_norm": 0.87890625,
      "learning_rate": 1.4489899970826085e-05,
      "loss": 0.0048,
      "step": 2522
    },
    {
      "epoch": 38.815384615384616,
      "grad_norm": 1.0625,
      "learning_rate": 1.4481284125651162e-05,
      "loss": 0.0041,
      "step": 2523
    },
    {
      "epoch": 38.83076923076923,
      "grad_norm": 0.75390625,
      "learning_rate": 1.4472667314822425e-05,
      "loss": 0.0035,
      "step": 2524
    },
    {
      "epoch": 38.84615384615385,
      "grad_norm": 2.703125,
      "learning_rate": 1.4464049542539598e-05,
      "loss": 0.0223,
      "step": 2525
    },
    {
      "epoch": 38.86153846153846,
      "grad_norm": 2.8125,
      "learning_rate": 1.4455430813002888e-05,
      "loss": 0.0175,
      "step": 2526
    },
    {
      "epoch": 38.87692307692308,
      "grad_norm": 0.765625,
      "learning_rate": 1.4446811130412952e-05,
      "loss": 0.0237,
      "step": 2527
    },
    {
      "epoch": 38.89230769230769,
      "grad_norm": 6.34375,
      "learning_rate": 1.4438190498970924e-05,
      "loss": 0.0303,
      "step": 2528
    },
    {
      "epoch": 38.90769230769231,
      "grad_norm": 3.09375,
      "learning_rate": 1.4429568922878392e-05,
      "loss": 0.0359,
      "step": 2529
    },
    {
      "epoch": 38.92307692307692,
      "grad_norm": 0.97265625,
      "learning_rate": 1.4420946406337404e-05,
      "loss": 0.0089,
      "step": 2530
    },
    {
      "epoch": 38.93846153846154,
      "grad_norm": 0.2119140625,
      "learning_rate": 1.4412322953550477e-05,
      "loss": 0.0128,
      "step": 2531
    },
    {
      "epoch": 38.95384615384615,
      "grad_norm": 1.34375,
      "learning_rate": 1.4403698568720572e-05,
      "loss": 0.0404,
      "step": 2532
    },
    {
      "epoch": 38.96923076923077,
      "grad_norm": 8.6875,
      "learning_rate": 1.4395073256051114e-05,
      "loss": 0.0675,
      "step": 2533
    },
    {
      "epoch": 38.98461538461538,
      "grad_norm": 0.13671875,
      "learning_rate": 1.438644701974597e-05,
      "loss": 0.0069,
      "step": 2534
    },
    {
      "epoch": 39.0,
      "grad_norm": 2.171875,
      "learning_rate": 1.4377819864009465e-05,
      "loss": 0.0094,
      "step": 2535
    },
    {
      "epoch": 39.01538461538462,
      "grad_norm": 4.0625,
      "learning_rate": 1.4369191793046372e-05,
      "loss": 0.0495,
      "step": 2536
    },
    {
      "epoch": 39.03076923076923,
      "grad_norm": 0.10595703125,
      "learning_rate": 1.4360562811061903e-05,
      "loss": 0.0134,
      "step": 2537
    },
    {
      "epoch": 39.04615384615385,
      "grad_norm": 2.03125,
      "learning_rate": 1.4351932922261726e-05,
      "loss": 0.0179,
      "step": 2538
    },
    {
      "epoch": 39.06153846153846,
      "grad_norm": 7.78125,
      "learning_rate": 1.4343302130851938e-05,
      "loss": 0.0332,
      "step": 2539
    },
    {
      "epoch": 39.07692307692308,
      "grad_norm": 1.8125,
      "learning_rate": 1.4334670441039083e-05,
      "loss": 0.0092,
      "step": 2540
    },
    {
      "epoch": 39.09230769230769,
      "grad_norm": 4.0,
      "learning_rate": 1.4326037857030148e-05,
      "loss": 0.0194,
      "step": 2541
    },
    {
      "epoch": 39.10769230769231,
      "grad_norm": 0.6953125,
      "learning_rate": 1.431740438303254e-05,
      "loss": 0.034,
      "step": 2542
    },
    {
      "epoch": 39.12307692307692,
      "grad_norm": 0.26171875,
      "learning_rate": 1.4308770023254114e-05,
      "loss": 0.0186,
      "step": 2543
    },
    {
      "epoch": 39.13846153846154,
      "grad_norm": 3.15625,
      "learning_rate": 1.4300134781903151e-05,
      "loss": 0.0278,
      "step": 2544
    },
    {
      "epoch": 39.15384615384615,
      "grad_norm": 0.1162109375,
      "learning_rate": 1.429149866318837e-05,
      "loss": 0.0296,
      "step": 2545
    },
    {
      "epoch": 39.16923076923077,
      "grad_norm": 2.203125,
      "learning_rate": 1.4282861671318898e-05,
      "loss": 0.0422,
      "step": 2546
    },
    {
      "epoch": 39.184615384615384,
      "grad_norm": 2.15625,
      "learning_rate": 1.4274223810504305e-05,
      "loss": 0.0177,
      "step": 2547
    },
    {
      "epoch": 39.2,
      "grad_norm": 2.65625,
      "learning_rate": 1.4265585084954586e-05,
      "loss": 0.0382,
      "step": 2548
    },
    {
      "epoch": 39.215384615384615,
      "grad_norm": 4.625,
      "learning_rate": 1.4256945498880136e-05,
      "loss": 0.0364,
      "step": 2549
    },
    {
      "epoch": 39.23076923076923,
      "grad_norm": 1.484375,
      "learning_rate": 1.4248305056491804e-05,
      "loss": 0.0107,
      "step": 2550
    },
    {
      "epoch": 39.246153846153845,
      "grad_norm": 0.59765625,
      "learning_rate": 1.4239663762000818e-05,
      "loss": 0.0322,
      "step": 2551
    },
    {
      "epoch": 39.261538461538464,
      "grad_norm": 3.703125,
      "learning_rate": 1.4231021619618854e-05,
      "loss": 0.0218,
      "step": 2552
    },
    {
      "epoch": 39.276923076923076,
      "grad_norm": 0.16015625,
      "learning_rate": 1.4222378633557976e-05,
      "loss": 0.0076,
      "step": 2553
    },
    {
      "epoch": 39.292307692307695,
      "grad_norm": 4.40625,
      "learning_rate": 1.4213734808030682e-05,
      "loss": 0.0272,
      "step": 2554
    },
    {
      "epoch": 39.30769230769231,
      "grad_norm": 10.4375,
      "learning_rate": 1.4205090147249856e-05,
      "loss": 0.0576,
      "step": 2555
    },
    {
      "epoch": 39.323076923076925,
      "grad_norm": 1.3828125,
      "learning_rate": 1.4196444655428806e-05,
      "loss": 0.0154,
      "step": 2556
    },
    {
      "epoch": 39.33846153846154,
      "grad_norm": 6.375,
      "learning_rate": 1.4187798336781241e-05,
      "loss": 0.0203,
      "step": 2557
    },
    {
      "epoch": 39.353846153846156,
      "grad_norm": 4.4375,
      "learning_rate": 1.4179151195521268e-05,
      "loss": 0.0228,
      "step": 2558
    },
    {
      "epoch": 39.36923076923077,
      "grad_norm": 1.90625,
      "learning_rate": 1.4170503235863408e-05,
      "loss": 0.0131,
      "step": 2559
    },
    {
      "epoch": 39.38461538461539,
      "grad_norm": 0.76953125,
      "learning_rate": 1.4161854462022556e-05,
      "loss": 0.0057,
      "step": 2560
    },
    {
      "epoch": 39.4,
      "grad_norm": 0.48046875,
      "learning_rate": 1.4153204878214032e-05,
      "loss": 0.0045,
      "step": 2561
    },
    {
      "epoch": 39.41538461538462,
      "grad_norm": 2.5,
      "learning_rate": 1.4144554488653533e-05,
      "loss": 0.0406,
      "step": 2562
    },
    {
      "epoch": 39.43076923076923,
      "grad_norm": 1.6484375,
      "learning_rate": 1.4135903297557155e-05,
      "loss": 0.0492,
      "step": 2563
    },
    {
      "epoch": 39.44615384615385,
      "grad_norm": 4.875,
      "learning_rate": 1.4127251309141381e-05,
      "loss": 0.0597,
      "step": 2564
    },
    {
      "epoch": 39.46153846153846,
      "grad_norm": 2.21875,
      "learning_rate": 1.4118598527623086e-05,
      "loss": 0.0621,
      "step": 2565
    },
    {
      "epoch": 39.47692307692308,
      "grad_norm": 4.09375,
      "learning_rate": 1.4109944957219537e-05,
      "loss": 0.0269,
      "step": 2566
    },
    {
      "epoch": 39.49230769230769,
      "grad_norm": 0.4140625,
      "learning_rate": 1.4101290602148371e-05,
      "loss": 0.0097,
      "step": 2567
    },
    {
      "epoch": 39.50769230769231,
      "grad_norm": 2.328125,
      "learning_rate": 1.4092635466627618e-05,
      "loss": 0.0186,
      "step": 2568
    },
    {
      "epoch": 39.52307692307692,
      "grad_norm": 1.84375,
      "learning_rate": 1.4083979554875685e-05,
      "loss": 0.0366,
      "step": 2569
    },
    {
      "epoch": 39.53846153846154,
      "grad_norm": 1.625,
      "learning_rate": 1.4075322871111357e-05,
      "loss": 0.0101,
      "step": 2570
    },
    {
      "epoch": 39.55384615384615,
      "grad_norm": 0.38671875,
      "learning_rate": 1.4066665419553804e-05,
      "loss": 0.0197,
      "step": 2571
    },
    {
      "epoch": 39.56923076923077,
      "grad_norm": 10.6875,
      "learning_rate": 1.4058007204422558e-05,
      "loss": 0.0634,
      "step": 2572
    },
    {
      "epoch": 39.58461538461538,
      "grad_norm": 9.0625,
      "learning_rate": 1.4049348229937523e-05,
      "loss": 0.0544,
      "step": 2573
    },
    {
      "epoch": 39.6,
      "grad_norm": 1.421875,
      "learning_rate": 1.4040688500318985e-05,
      "loss": 0.0088,
      "step": 2574
    },
    {
      "epoch": 39.61538461538461,
      "grad_norm": 2.984375,
      "learning_rate": 1.4032028019787585e-05,
      "loss": 0.0273,
      "step": 2575
    },
    {
      "epoch": 39.63076923076923,
      "grad_norm": 1.1328125,
      "learning_rate": 1.4023366792564343e-05,
      "loss": 0.0151,
      "step": 2576
    },
    {
      "epoch": 39.646153846153844,
      "grad_norm": 0.29296875,
      "learning_rate": 1.4014704822870634e-05,
      "loss": 0.0094,
      "step": 2577
    },
    {
      "epoch": 39.66153846153846,
      "grad_norm": 1.90625,
      "learning_rate": 1.4006042114928191e-05,
      "loss": 0.0234,
      "step": 2578
    },
    {
      "epoch": 39.676923076923075,
      "grad_norm": 1.015625,
      "learning_rate": 1.3997378672959122e-05,
      "loss": 0.0112,
      "step": 2579
    },
    {
      "epoch": 39.69230769230769,
      "grad_norm": 1.625,
      "learning_rate": 1.3988714501185873e-05,
      "loss": 0.0091,
      "step": 2580
    },
    {
      "epoch": 39.707692307692305,
      "grad_norm": 2.90625,
      "learning_rate": 1.3980049603831266e-05,
      "loss": 0.0432,
      "step": 2581
    },
    {
      "epoch": 39.723076923076924,
      "grad_norm": 0.34765625,
      "learning_rate": 1.3971383985118455e-05,
      "loss": 0.0027,
      "step": 2582
    },
    {
      "epoch": 39.738461538461536,
      "grad_norm": 1.6953125,
      "learning_rate": 1.396271764927097e-05,
      "loss": 0.0316,
      "step": 2583
    },
    {
      "epoch": 39.753846153846155,
      "grad_norm": 0.4140625,
      "learning_rate": 1.3954050600512669e-05,
      "loss": 0.0048,
      "step": 2584
    },
    {
      "epoch": 39.76923076923077,
      "grad_norm": 0.76953125,
      "learning_rate": 1.3945382843067772e-05,
      "loss": 0.024,
      "step": 2585
    },
    {
      "epoch": 39.784615384615385,
      "grad_norm": 1.4375,
      "learning_rate": 1.3936714381160835e-05,
      "loss": 0.0162,
      "step": 2586
    },
    {
      "epoch": 39.8,
      "grad_norm": 3.28125,
      "learning_rate": 1.3928045219016761e-05,
      "loss": 0.0356,
      "step": 2587
    },
    {
      "epoch": 39.815384615384616,
      "grad_norm": 2.34375,
      "learning_rate": 1.3919375360860795e-05,
      "loss": 0.0143,
      "step": 2588
    },
    {
      "epoch": 39.83076923076923,
      "grad_norm": 2.296875,
      "learning_rate": 1.391070481091852e-05,
      "loss": 0.0138,
      "step": 2589
    },
    {
      "epoch": 39.84615384615385,
      "grad_norm": 2.59375,
      "learning_rate": 1.3902033573415862e-05,
      "loss": 0.0195,
      "step": 2590
    },
    {
      "epoch": 39.86153846153846,
      "grad_norm": 5.1875,
      "learning_rate": 1.3893361652579068e-05,
      "loss": 0.0472,
      "step": 2591
    },
    {
      "epoch": 39.87692307692308,
      "grad_norm": 2.234375,
      "learning_rate": 1.3884689052634733e-05,
      "loss": 0.0522,
      "step": 2592
    },
    {
      "epoch": 39.89230769230769,
      "grad_norm": 2.203125,
      "learning_rate": 1.3876015777809778e-05,
      "loss": 0.0105,
      "step": 2593
    },
    {
      "epoch": 39.90769230769231,
      "grad_norm": 1.0625,
      "learning_rate": 1.3867341832331446e-05,
      "loss": 0.0135,
      "step": 2594
    },
    {
      "epoch": 39.92307692307692,
      "grad_norm": 1.8515625,
      "learning_rate": 1.3858667220427315e-05,
      "loss": 0.0179,
      "step": 2595
    },
    {
      "epoch": 39.93846153846154,
      "grad_norm": 3.328125,
      "learning_rate": 1.3849991946325287e-05,
      "loss": 0.0162,
      "step": 2596
    },
    {
      "epoch": 39.95384615384615,
      "grad_norm": 8.375,
      "learning_rate": 1.3841316014253584e-05,
      "loss": 0.037,
      "step": 2597
    },
    {
      "epoch": 39.96923076923077,
      "grad_norm": 2.296875,
      "learning_rate": 1.383263942844075e-05,
      "loss": 0.0224,
      "step": 2598
    },
    {
      "epoch": 39.98461538461538,
      "grad_norm": 1.375,
      "learning_rate": 1.382396219311565e-05,
      "loss": 0.0202,
      "step": 2599
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.83203125,
      "learning_rate": 1.3815284312507459e-05,
      "loss": 0.0174,
      "step": 2600
    },
    {
      "epoch": 40.01538461538462,
      "grad_norm": 1.9765625,
      "learning_rate": 1.380660579084567e-05,
      "loss": 0.0052,
      "step": 2601
    },
    {
      "epoch": 40.03076923076923,
      "grad_norm": 1.234375,
      "learning_rate": 1.3797926632360092e-05,
      "loss": 0.0186,
      "step": 2602
    },
    {
      "epoch": 40.04615384615385,
      "grad_norm": 0.130859375,
      "learning_rate": 1.3789246841280842e-05,
      "loss": 0.0192,
      "step": 2603
    },
    {
      "epoch": 40.06153846153846,
      "grad_norm": 0.9609375,
      "learning_rate": 1.3780566421838346e-05,
      "loss": 0.0102,
      "step": 2604
    },
    {
      "epoch": 40.07692307692308,
      "grad_norm": 1.734375,
      "learning_rate": 1.377188537826333e-05,
      "loss": 0.0054,
      "step": 2605
    },
    {
      "epoch": 40.09230769230769,
      "grad_norm": 1.9140625,
      "learning_rate": 1.3763203714786829e-05,
      "loss": 0.031,
      "step": 2606
    },
    {
      "epoch": 40.10769230769231,
      "grad_norm": 0.125,
      "learning_rate": 1.3754521435640186e-05,
      "loss": 0.0074,
      "step": 2607
    },
    {
      "epoch": 40.12307692307692,
      "grad_norm": 0.400390625,
      "learning_rate": 1.3745838545055038e-05,
      "loss": 0.0026,
      "step": 2608
    },
    {
      "epoch": 40.13846153846154,
      "grad_norm": 2.515625,
      "learning_rate": 1.3737155047263315e-05,
      "loss": 0.0242,
      "step": 2609
    },
    {
      "epoch": 40.15384615384615,
      "grad_norm": 0.9453125,
      "learning_rate": 1.3728470946497248e-05,
      "loss": 0.0117,
      "step": 2610
    },
    {
      "epoch": 40.16923076923077,
      "grad_norm": 4.4375,
      "learning_rate": 1.3719786246989368e-05,
      "loss": 0.0498,
      "step": 2611
    },
    {
      "epoch": 40.184615384615384,
      "grad_norm": 0.67578125,
      "learning_rate": 1.3711100952972491e-05,
      "loss": 0.0191,
      "step": 2612
    },
    {
      "epoch": 40.2,
      "grad_norm": 0.2294921875,
      "learning_rate": 1.3702415068679722e-05,
      "loss": 0.0095,
      "step": 2613
    },
    {
      "epoch": 40.215384615384615,
      "grad_norm": 1.4609375,
      "learning_rate": 1.3693728598344454e-05,
      "loss": 0.0207,
      "step": 2614
    },
    {
      "epoch": 40.23076923076923,
      "grad_norm": 3.28125,
      "learning_rate": 1.3685041546200367e-05,
      "loss": 0.0505,
      "step": 2615
    },
    {
      "epoch": 40.246153846153845,
      "grad_norm": 5.28125,
      "learning_rate": 1.3676353916481433e-05,
      "loss": 0.0331,
      "step": 2616
    },
    {
      "epoch": 40.261538461538464,
      "grad_norm": 1.046875,
      "learning_rate": 1.3667665713421884e-05,
      "loss": 0.0111,
      "step": 2617
    },
    {
      "epoch": 40.276923076923076,
      "grad_norm": 1.453125,
      "learning_rate": 1.3658976941256252e-05,
      "loss": 0.0224,
      "step": 2618
    },
    {
      "epoch": 40.292307692307695,
      "grad_norm": 4.0625,
      "learning_rate": 1.3650287604219342e-05,
      "loss": 0.0247,
      "step": 2619
    },
    {
      "epoch": 40.30769230769231,
      "grad_norm": 3.59375,
      "learning_rate": 1.3641597706546223e-05,
      "loss": 0.0382,
      "step": 2620
    },
    {
      "epoch": 40.323076923076925,
      "grad_norm": 4.125,
      "learning_rate": 1.3632907252472255e-05,
      "loss": 0.0323,
      "step": 2621
    },
    {
      "epoch": 40.33846153846154,
      "grad_norm": 2.453125,
      "learning_rate": 1.3624216246233048e-05,
      "loss": 0.0409,
      "step": 2622
    },
    {
      "epoch": 40.353846153846156,
      "grad_norm": 2.484375,
      "learning_rate": 1.3615524692064508e-05,
      "loss": 0.0132,
      "step": 2623
    },
    {
      "epoch": 40.36923076923077,
      "grad_norm": 0.2392578125,
      "learning_rate": 1.3606832594202778e-05,
      "loss": 0.005,
      "step": 2624
    },
    {
      "epoch": 40.38461538461539,
      "grad_norm": 1.78125,
      "learning_rate": 1.3598139956884292e-05,
      "loss": 0.0343,
      "step": 2625
    },
    {
      "epoch": 40.4,
      "grad_norm": 0.4375,
      "learning_rate": 1.358944678434573e-05,
      "loss": 0.0224,
      "step": 2626
    },
    {
      "epoch": 40.41538461538462,
      "grad_norm": 0.94921875,
      "learning_rate": 1.3580753080824038e-05,
      "loss": 0.0147,
      "step": 2627
    },
    {
      "epoch": 40.43076923076923,
      "grad_norm": 0.65234375,
      "learning_rate": 1.3572058850556423e-05,
      "loss": 0.0182,
      "step": 2628
    },
    {
      "epoch": 40.44615384615385,
      "grad_norm": 2.75,
      "learning_rate": 1.3563364097780351e-05,
      "loss": 0.0124,
      "step": 2629
    },
    {
      "epoch": 40.46153846153846,
      "grad_norm": 4.8125,
      "learning_rate": 1.3554668826733535e-05,
      "loss": 0.0587,
      "step": 2630
    },
    {
      "epoch": 40.47692307692308,
      "grad_norm": 0.16796875,
      "learning_rate": 1.3545973041653948e-05,
      "loss": 0.0119,
      "step": 2631
    },
    {
      "epoch": 40.49230769230769,
      "grad_norm": 5.0625,
      "learning_rate": 1.35372767467798e-05,
      "loss": 0.067,
      "step": 2632
    },
    {
      "epoch": 40.50769230769231,
      "grad_norm": 3.5,
      "learning_rate": 1.352857994634957e-05,
      "loss": 0.0503,
      "step": 2633
    },
    {
      "epoch": 40.52307692307692,
      "grad_norm": 1.6484375,
      "learning_rate": 1.3519882644601973e-05,
      "loss": 0.0234,
      "step": 2634
    },
    {
      "epoch": 40.53846153846154,
      "grad_norm": 2.90625,
      "learning_rate": 1.3511184845775959e-05,
      "loss": 0.0664,
      "step": 2635
    },
    {
      "epoch": 40.55384615384615,
      "grad_norm": 3.09375,
      "learning_rate": 1.3502486554110736e-05,
      "loss": 0.0161,
      "step": 2636
    },
    {
      "epoch": 40.56923076923077,
      "grad_norm": 0.50390625,
      "learning_rate": 1.3493787773845749e-05,
      "loss": 0.0034,
      "step": 2637
    },
    {
      "epoch": 40.58461538461538,
      "grad_norm": 0.7421875,
      "learning_rate": 1.348508850922067e-05,
      "loss": 0.01,
      "step": 2638
    },
    {
      "epoch": 40.6,
      "grad_norm": 2.71875,
      "learning_rate": 1.347638876447542e-05,
      "loss": 0.0436,
      "step": 2639
    },
    {
      "epoch": 40.61538461538461,
      "grad_norm": 0.87109375,
      "learning_rate": 1.3467688543850149e-05,
      "loss": 0.0099,
      "step": 2640
    },
    {
      "epoch": 40.63076923076923,
      "grad_norm": 2.015625,
      "learning_rate": 1.3458987851585233e-05,
      "loss": 0.0125,
      "step": 2641
    },
    {
      "epoch": 40.646153846153844,
      "grad_norm": 1.7265625,
      "learning_rate": 1.3450286691921291e-05,
      "loss": 0.0084,
      "step": 2642
    },
    {
      "epoch": 40.66153846153846,
      "grad_norm": 0.5546875,
      "learning_rate": 1.3441585069099161e-05,
      "loss": 0.0109,
      "step": 2643
    },
    {
      "epoch": 40.676923076923075,
      "grad_norm": 7.5,
      "learning_rate": 1.3432882987359905e-05,
      "loss": 0.0389,
      "step": 2644
    },
    {
      "epoch": 40.69230769230769,
      "grad_norm": 2.71875,
      "learning_rate": 1.3424180450944815e-05,
      "loss": 0.0205,
      "step": 2645
    },
    {
      "epoch": 40.707692307692305,
      "grad_norm": 0.21484375,
      "learning_rate": 1.3415477464095397e-05,
      "loss": 0.041,
      "step": 2646
    },
    {
      "epoch": 40.723076923076924,
      "grad_norm": 1.4765625,
      "learning_rate": 1.3406774031053387e-05,
      "loss": 0.0237,
      "step": 2647
    },
    {
      "epoch": 40.738461538461536,
      "grad_norm": 2.484375,
      "learning_rate": 1.3398070156060729e-05,
      "loss": 0.0204,
      "step": 2648
    },
    {
      "epoch": 40.753846153846155,
      "grad_norm": 0.64453125,
      "learning_rate": 1.3389365843359589e-05,
      "loss": 0.0138,
      "step": 2649
    },
    {
      "epoch": 40.76923076923077,
      "grad_norm": 3.53125,
      "learning_rate": 1.3380661097192337e-05,
      "loss": 0.0239,
      "step": 2650
    },
    {
      "epoch": 40.784615384615385,
      "grad_norm": 5.28125,
      "learning_rate": 1.3371955921801566e-05,
      "loss": 0.0136,
      "step": 2651
    },
    {
      "epoch": 40.8,
      "grad_norm": 0.240234375,
      "learning_rate": 1.3363250321430077e-05,
      "loss": 0.004,
      "step": 2652
    },
    {
      "epoch": 40.815384615384616,
      "grad_norm": 4.59375,
      "learning_rate": 1.3354544300320859e-05,
      "loss": 0.0347,
      "step": 2653
    },
    {
      "epoch": 40.83076923076923,
      "grad_norm": 1.1875,
      "learning_rate": 1.3345837862717136e-05,
      "loss": 0.0138,
      "step": 2654
    },
    {
      "epoch": 40.84615384615385,
      "grad_norm": 1.8203125,
      "learning_rate": 1.3337131012862313e-05,
      "loss": 0.012,
      "step": 2655
    },
    {
      "epoch": 40.86153846153846,
      "grad_norm": 0.494140625,
      "learning_rate": 1.3328423755000006e-05,
      "loss": 0.0055,
      "step": 2656
    },
    {
      "epoch": 40.87692307692308,
      "grad_norm": 0.84375,
      "learning_rate": 1.3319716093374021e-05,
      "loss": 0.0305,
      "step": 2657
    },
    {
      "epoch": 40.89230769230769,
      "grad_norm": 1.359375,
      "learning_rate": 1.3311008032228369e-05,
      "loss": 0.008,
      "step": 2658
    },
    {
      "epoch": 40.90769230769231,
      "grad_norm": 0.3828125,
      "learning_rate": 1.3302299575807257e-05,
      "loss": 0.0329,
      "step": 2659
    },
    {
      "epoch": 40.92307692307692,
      "grad_norm": 1.1953125,
      "learning_rate": 1.3293590728355077e-05,
      "loss": 0.0146,
      "step": 2660
    },
    {
      "epoch": 40.93846153846154,
      "grad_norm": 2.796875,
      "learning_rate": 1.328488149411642e-05,
      "loss": 0.0232,
      "step": 2661
    },
    {
      "epoch": 40.95384615384615,
      "grad_norm": 0.404296875,
      "learning_rate": 1.3276171877336051e-05,
      "loss": 0.0242,
      "step": 2662
    },
    {
      "epoch": 40.96923076923077,
      "grad_norm": 5.375,
      "learning_rate": 1.3267461882258945e-05,
      "loss": 0.0288,
      "step": 2663
    },
    {
      "epoch": 40.98461538461538,
      "grad_norm": 1.2734375,
      "learning_rate": 1.3258751513130239e-05,
      "loss": 0.0292,
      "step": 2664
    },
    {
      "epoch": 41.0,
      "grad_norm": 2.53125,
      "learning_rate": 1.325004077419526e-05,
      "loss": 0.0154,
      "step": 2665
    },
    {
      "epoch": 41.01538461538462,
      "grad_norm": 6.59375,
      "learning_rate": 1.3241329669699523e-05,
      "loss": 0.0279,
      "step": 2666
    },
    {
      "epoch": 41.03076923076923,
      "grad_norm": 0.162109375,
      "learning_rate": 1.3232618203888705e-05,
      "loss": 0.0149,
      "step": 2667
    },
    {
      "epoch": 41.04615384615385,
      "grad_norm": 4.46875,
      "learning_rate": 1.322390638100868e-05,
      "loss": 0.021,
      "step": 2668
    },
    {
      "epoch": 41.06153846153846,
      "grad_norm": 1.203125,
      "learning_rate": 1.3215194205305478e-05,
      "loss": 0.0057,
      "step": 2669
    },
    {
      "epoch": 41.07692307692308,
      "grad_norm": 2.140625,
      "learning_rate": 1.3206481681025309e-05,
      "loss": 0.0266,
      "step": 2670
    },
    {
      "epoch": 41.09230769230769,
      "grad_norm": 0.51953125,
      "learning_rate": 1.3197768812414548e-05,
      "loss": 0.0113,
      "step": 2671
    },
    {
      "epoch": 41.10769230769231,
      "grad_norm": 0.515625,
      "learning_rate": 1.3189055603719747e-05,
      "loss": 0.0142,
      "step": 2672
    },
    {
      "epoch": 41.12307692307692,
      "grad_norm": 0.6328125,
      "learning_rate": 1.3180342059187614e-05,
      "loss": 0.0044,
      "step": 2673
    },
    {
      "epoch": 41.13846153846154,
      "grad_norm": 4.0625,
      "learning_rate": 1.3171628183065027e-05,
      "loss": 0.0301,
      "step": 2674
    },
    {
      "epoch": 41.15384615384615,
      "grad_norm": 0.3515625,
      "learning_rate": 1.3162913979599022e-05,
      "loss": 0.0113,
      "step": 2675
    },
    {
      "epoch": 41.16923076923077,
      "grad_norm": 1.234375,
      "learning_rate": 1.31541994530368e-05,
      "loss": 0.007,
      "step": 2676
    },
    {
      "epoch": 41.184615384615384,
      "grad_norm": 2.078125,
      "learning_rate": 1.3145484607625708e-05,
      "loss": 0.0133,
      "step": 2677
    },
    {
      "epoch": 41.2,
      "grad_norm": 0.56640625,
      "learning_rate": 1.3136769447613261e-05,
      "loss": 0.0106,
      "step": 2678
    },
    {
      "epoch": 41.215384615384615,
      "grad_norm": 1.265625,
      "learning_rate": 1.3128053977247121e-05,
      "loss": 0.0099,
      "step": 2679
    },
    {
      "epoch": 41.23076923076923,
      "grad_norm": 4.96875,
      "learning_rate": 1.3119338200775103e-05,
      "loss": 0.0257,
      "step": 2680
    },
    {
      "epoch": 41.246153846153845,
      "grad_norm": 2.359375,
      "learning_rate": 1.3110622122445168e-05,
      "loss": 0.0324,
      "step": 2681
    },
    {
      "epoch": 41.261538461538464,
      "grad_norm": 1.0390625,
      "learning_rate": 1.3101905746505428e-05,
      "loss": 0.017,
      "step": 2682
    },
    {
      "epoch": 41.276923076923076,
      "grad_norm": 0.474609375,
      "learning_rate": 1.3093189077204143e-05,
      "loss": 0.0033,
      "step": 2683
    },
    {
      "epoch": 41.292307692307695,
      "grad_norm": 1.8125,
      "learning_rate": 1.3084472118789703e-05,
      "loss": 0.021,
      "step": 2684
    },
    {
      "epoch": 41.30769230769231,
      "grad_norm": 3.1875,
      "learning_rate": 1.3075754875510654e-05,
      "loss": 0.0255,
      "step": 2685
    },
    {
      "epoch": 41.323076923076925,
      "grad_norm": 2.03125,
      "learning_rate": 1.3067037351615672e-05,
      "loss": 0.0128,
      "step": 2686
    },
    {
      "epoch": 41.33846153846154,
      "grad_norm": 3.59375,
      "learning_rate": 1.3058319551353578e-05,
      "loss": 0.0161,
      "step": 2687
    },
    {
      "epoch": 41.353846153846156,
      "grad_norm": 3.6875,
      "learning_rate": 1.3049601478973312e-05,
      "loss": 0.0318,
      "step": 2688
    },
    {
      "epoch": 41.36923076923077,
      "grad_norm": 0.78125,
      "learning_rate": 1.3040883138723964e-05,
      "loss": 0.0063,
      "step": 2689
    },
    {
      "epoch": 41.38461538461539,
      "grad_norm": 0.546875,
      "learning_rate": 1.3032164534854747e-05,
      "loss": 0.0101,
      "step": 2690
    },
    {
      "epoch": 41.4,
      "grad_norm": 0.35546875,
      "learning_rate": 1.3023445671614995e-05,
      "loss": 0.0084,
      "step": 2691
    },
    {
      "epoch": 41.41538461538462,
      "grad_norm": 1.2734375,
      "learning_rate": 1.301472655325419e-05,
      "loss": 0.0586,
      "step": 2692
    },
    {
      "epoch": 41.43076923076923,
      "grad_norm": 2.4375,
      "learning_rate": 1.3006007184021907e-05,
      "loss": 0.0211,
      "step": 2693
    },
    {
      "epoch": 41.44615384615385,
      "grad_norm": 5.1875,
      "learning_rate": 1.2997287568167874e-05,
      "loss": 0.0665,
      "step": 2694
    },
    {
      "epoch": 41.46153846153846,
      "grad_norm": 3.53125,
      "learning_rate": 1.2988567709941921e-05,
      "loss": 0.0441,
      "step": 2695
    },
    {
      "epoch": 41.47692307692308,
      "grad_norm": 3.34375,
      "learning_rate": 1.2979847613594002e-05,
      "loss": 0.0226,
      "step": 2696
    },
    {
      "epoch": 41.49230769230769,
      "grad_norm": 6.78125,
      "learning_rate": 1.2971127283374185e-05,
      "loss": 0.0451,
      "step": 2697
    },
    {
      "epoch": 41.50769230769231,
      "grad_norm": 6.375,
      "learning_rate": 1.296240672353265e-05,
      "loss": 0.025,
      "step": 2698
    },
    {
      "epoch": 41.52307692307692,
      "grad_norm": 1.3046875,
      "learning_rate": 1.2953685938319699e-05,
      "loss": 0.0163,
      "step": 2699
    },
    {
      "epoch": 41.53846153846154,
      "grad_norm": 2.8125,
      "learning_rate": 1.2944964931985731e-05,
      "loss": 0.0169,
      "step": 2700
    },
    {
      "epoch": 41.55384615384615,
      "grad_norm": 1.46875,
      "learning_rate": 1.2936243708781266e-05,
      "loss": 0.0082,
      "step": 2701
    },
    {
      "epoch": 41.56923076923077,
      "grad_norm": 6.8125,
      "learning_rate": 1.2927522272956912e-05,
      "loss": 0.057,
      "step": 2702
    },
    {
      "epoch": 41.58461538461538,
      "grad_norm": 0.703125,
      "learning_rate": 1.2918800628763394e-05,
      "loss": 0.0061,
      "step": 2703
    },
    {
      "epoch": 41.6,
      "grad_norm": 1.8203125,
      "learning_rate": 1.291007878045154e-05,
      "loss": 0.0245,
      "step": 2704
    },
    {
      "epoch": 41.61538461538461,
      "grad_norm": 0.8671875,
      "learning_rate": 1.2901356732272271e-05,
      "loss": 0.0068,
      "step": 2705
    },
    {
      "epoch": 41.63076923076923,
      "grad_norm": 0.96484375,
      "learning_rate": 1.2892634488476605e-05,
      "loss": 0.0196,
      "step": 2706
    },
    {
      "epoch": 41.646153846153844,
      "grad_norm": 1.21875,
      "learning_rate": 1.288391205331566e-05,
      "loss": 0.0229,
      "step": 2707
    },
    {
      "epoch": 41.66153846153846,
      "grad_norm": 0.49609375,
      "learning_rate": 1.287518943104064e-05,
      "loss": 0.0077,
      "step": 2708
    },
    {
      "epoch": 41.676923076923075,
      "grad_norm": 2.6875,
      "learning_rate": 1.2866466625902857e-05,
      "loss": 0.0177,
      "step": 2709
    },
    {
      "epoch": 41.69230769230769,
      "grad_norm": 2.25,
      "learning_rate": 1.2857743642153685e-05,
      "loss": 0.0169,
      "step": 2710
    },
    {
      "epoch": 41.707692307692305,
      "grad_norm": 4.5625,
      "learning_rate": 1.2849020484044613e-05,
      "loss": 0.0266,
      "step": 2711
    },
    {
      "epoch": 41.723076923076924,
      "grad_norm": 3.78125,
      "learning_rate": 1.284029715582719e-05,
      "loss": 0.0198,
      "step": 2712
    },
    {
      "epoch": 41.738461538461536,
      "grad_norm": 3.421875,
      "learning_rate": 1.2831573661753074e-05,
      "loss": 0.0241,
      "step": 2713
    },
    {
      "epoch": 41.753846153846155,
      "grad_norm": 5.65625,
      "learning_rate": 1.2822850006073986e-05,
      "loss": 0.0376,
      "step": 2714
    },
    {
      "epoch": 41.76923076923077,
      "grad_norm": 0.3046875,
      "learning_rate": 1.281412619304172e-05,
      "loss": 0.0366,
      "step": 2715
    },
    {
      "epoch": 41.784615384615385,
      "grad_norm": 0.921875,
      "learning_rate": 1.2805402226908166e-05,
      "loss": 0.0206,
      "step": 2716
    },
    {
      "epoch": 41.8,
      "grad_norm": 1.359375,
      "learning_rate": 1.2796678111925276e-05,
      "loss": 0.0072,
      "step": 2717
    },
    {
      "epoch": 41.815384615384616,
      "grad_norm": 2.21875,
      "learning_rate": 1.278795385234508e-05,
      "loss": 0.078,
      "step": 2718
    },
    {
      "epoch": 41.83076923076923,
      "grad_norm": 1.859375,
      "learning_rate": 1.2779229452419673e-05,
      "loss": 0.0202,
      "step": 2719
    },
    {
      "epoch": 41.84615384615385,
      "grad_norm": 0.185546875,
      "learning_rate": 1.2770504916401224e-05,
      "loss": 0.0094,
      "step": 2720
    },
    {
      "epoch": 41.86153846153846,
      "grad_norm": 7.03125,
      "learning_rate": 1.2761780248541963e-05,
      "loss": 0.0322,
      "step": 2721
    },
    {
      "epoch": 41.87692307692308,
      "grad_norm": 4.25,
      "learning_rate": 1.275305545309419e-05,
      "loss": 0.0254,
      "step": 2722
    },
    {
      "epoch": 41.89230769230769,
      "grad_norm": 3.828125,
      "learning_rate": 1.2744330534310267e-05,
      "loss": 0.0454,
      "step": 2723
    },
    {
      "epoch": 41.90769230769231,
      "grad_norm": 2.296875,
      "learning_rate": 1.2735605496442604e-05,
      "loss": 0.0164,
      "step": 2724
    },
    {
      "epoch": 41.92307692307692,
      "grad_norm": 3.625,
      "learning_rate": 1.2726880343743685e-05,
      "loss": 0.0311,
      "step": 2725
    },
    {
      "epoch": 41.93846153846154,
      "grad_norm": 0.6015625,
      "learning_rate": 1.2718155080466046e-05,
      "loss": 0.0089,
      "step": 2726
    },
    {
      "epoch": 41.95384615384615,
      "grad_norm": 4.1875,
      "learning_rate": 1.2709429710862269e-05,
      "loss": 0.0308,
      "step": 2727
    },
    {
      "epoch": 41.96923076923077,
      "grad_norm": 2.1875,
      "learning_rate": 1.2700704239184995e-05,
      "loss": 0.0316,
      "step": 2728
    },
    {
      "epoch": 41.98461538461538,
      "grad_norm": 2.984375,
      "learning_rate": 1.2691978669686913e-05,
      "loss": 0.0149,
      "step": 2729
    },
    {
      "epoch": 42.0,
      "grad_norm": 0.5390625,
      "learning_rate": 1.268325300662076e-05,
      "loss": 0.0148,
      "step": 2730
    },
    {
      "epoch": 42.01538461538462,
      "grad_norm": 3.46875,
      "learning_rate": 1.2674527254239318e-05,
      "loss": 0.0493,
      "step": 2731
    },
    {
      "epoch": 42.03076923076923,
      "grad_norm": 2.59375,
      "learning_rate": 1.2665801416795412e-05,
      "loss": 0.0216,
      "step": 2732
    },
    {
      "epoch": 42.04615384615385,
      "grad_norm": 2.625,
      "learning_rate": 1.265707549854191e-05,
      "loss": 0.0222,
      "step": 2733
    },
    {
      "epoch": 42.06153846153846,
      "grad_norm": 2.25,
      "learning_rate": 1.2648349503731716e-05,
      "loss": 0.0344,
      "step": 2734
    },
    {
      "epoch": 42.07692307692308,
      "grad_norm": 3.78125,
      "learning_rate": 1.263962343661778e-05,
      "loss": 0.0203,
      "step": 2735
    },
    {
      "epoch": 42.09230769230769,
      "grad_norm": 0.97265625,
      "learning_rate": 1.2630897301453073e-05,
      "loss": 0.0059,
      "step": 2736
    },
    {
      "epoch": 42.10769230769231,
      "grad_norm": 1.34375,
      "learning_rate": 1.2622171102490615e-05,
      "loss": 0.0164,
      "step": 2737
    },
    {
      "epoch": 42.12307692307692,
      "grad_norm": 5.71875,
      "learning_rate": 1.2613444843983441e-05,
      "loss": 0.0337,
      "step": 2738
    },
    {
      "epoch": 42.13846153846154,
      "grad_norm": 1.453125,
      "learning_rate": 1.2604718530184634e-05,
      "loss": 0.0091,
      "step": 2739
    },
    {
      "epoch": 42.15384615384615,
      "grad_norm": 0.55859375,
      "learning_rate": 1.2595992165347287e-05,
      "loss": 0.0058,
      "step": 2740
    },
    {
      "epoch": 42.16923076923077,
      "grad_norm": 0.6875,
      "learning_rate": 1.258726575372452e-05,
      "loss": 0.0439,
      "step": 2741
    },
    {
      "epoch": 42.184615384615384,
      "grad_norm": 2.28125,
      "learning_rate": 1.2578539299569488e-05,
      "loss": 0.0205,
      "step": 2742
    },
    {
      "epoch": 42.2,
      "grad_norm": 1.46875,
      "learning_rate": 1.2569812807135353e-05,
      "loss": 0.0314,
      "step": 2743
    },
    {
      "epoch": 42.215384615384615,
      "grad_norm": 1.4296875,
      "learning_rate": 1.2561086280675308e-05,
      "loss": 0.0223,
      "step": 2744
    },
    {
      "epoch": 42.23076923076923,
      "grad_norm": 3.75,
      "learning_rate": 1.255235972444255e-05,
      "loss": 0.0128,
      "step": 2745
    },
    {
      "epoch": 42.246153846153845,
      "grad_norm": 0.29296875,
      "learning_rate": 1.2543633142690298e-05,
      "loss": 0.0116,
      "step": 2746
    },
    {
      "epoch": 42.261538461538464,
      "grad_norm": 2.9375,
      "learning_rate": 1.2534906539671785e-05,
      "loss": 0.0128,
      "step": 2747
    },
    {
      "epoch": 42.276923076923076,
      "grad_norm": 1.5546875,
      "learning_rate": 1.252617991964024e-05,
      "loss": 0.0379,
      "step": 2748
    },
    {
      "epoch": 42.292307692307695,
      "grad_norm": 2.84375,
      "learning_rate": 1.2517453286848932e-05,
      "loss": 0.0186,
      "step": 2749
    },
    {
      "epoch": 42.30769230769231,
      "grad_norm": 2.5625,
      "learning_rate": 1.2508726645551094e-05,
      "loss": 0.0373,
      "step": 2750
    },
    {
      "epoch": 42.323076923076925,
      "grad_norm": 9.25,
      "learning_rate": 1.25e-05,
      "loss": 0.0399,
      "step": 2751
    },
    {
      "epoch": 42.33846153846154,
      "grad_norm": 3.625,
      "learning_rate": 1.2491273354448907e-05,
      "loss": 0.036,
      "step": 2752
    },
    {
      "epoch": 42.353846153846156,
      "grad_norm": 3.515625,
      "learning_rate": 1.248254671315107e-05,
      "loss": 0.0308,
      "step": 2753
    },
    {
      "epoch": 42.36923076923077,
      "grad_norm": 2.234375,
      "learning_rate": 1.2473820080359758e-05,
      "loss": 0.0483,
      "step": 2754
    },
    {
      "epoch": 42.38461538461539,
      "grad_norm": 1.53125,
      "learning_rate": 1.246509346032822e-05,
      "loss": 0.0067,
      "step": 2755
    },
    {
      "epoch": 42.4,
      "grad_norm": 0.95703125,
      "learning_rate": 1.2456366857309709e-05,
      "loss": 0.0314,
      "step": 2756
    },
    {
      "epoch": 42.41538461538462,
      "grad_norm": 1.9765625,
      "learning_rate": 1.2447640275557453e-05,
      "loss": 0.026,
      "step": 2757
    },
    {
      "epoch": 42.43076923076923,
      "grad_norm": 1.9609375,
      "learning_rate": 1.2438913719324696e-05,
      "loss": 0.0181,
      "step": 2758
    },
    {
      "epoch": 42.44615384615385,
      "grad_norm": 1.1640625,
      "learning_rate": 1.243018719286465e-05,
      "loss": 0.0198,
      "step": 2759
    },
    {
      "epoch": 42.46153846153846,
      "grad_norm": 1.578125,
      "learning_rate": 1.2421460700430514e-05,
      "loss": 0.0298,
      "step": 2760
    },
    {
      "epoch": 42.47692307692308,
      "grad_norm": 4.78125,
      "learning_rate": 1.2412734246275482e-05,
      "loss": 0.0293,
      "step": 2761
    },
    {
      "epoch": 42.49230769230769,
      "grad_norm": 2.71875,
      "learning_rate": 1.2404007834652716e-05,
      "loss": 0.0162,
      "step": 2762
    },
    {
      "epoch": 42.50769230769231,
      "grad_norm": 1.5234375,
      "learning_rate": 1.2395281469815367e-05,
      "loss": 0.0357,
      "step": 2763
    },
    {
      "epoch": 42.52307692307692,
      "grad_norm": 2.8125,
      "learning_rate": 1.2386555156016558e-05,
      "loss": 0.0285,
      "step": 2764
    },
    {
      "epoch": 42.53846153846154,
      "grad_norm": 3.0625,
      "learning_rate": 1.237782889750939e-05,
      "loss": 0.0168,
      "step": 2765
    },
    {
      "epoch": 42.55384615384615,
      "grad_norm": 2.09375,
      "learning_rate": 1.2369102698546932e-05,
      "loss": 0.0255,
      "step": 2766
    },
    {
      "epoch": 42.56923076923077,
      "grad_norm": 4.25,
      "learning_rate": 1.2360376563382224e-05,
      "loss": 0.0257,
      "step": 2767
    },
    {
      "epoch": 42.58461538461538,
      "grad_norm": 2.265625,
      "learning_rate": 1.2351650496268287e-05,
      "loss": 0.026,
      "step": 2768
    },
    {
      "epoch": 42.6,
      "grad_norm": 1.640625,
      "learning_rate": 1.2342924501458091e-05,
      "loss": 0.0094,
      "step": 2769
    },
    {
      "epoch": 42.61538461538461,
      "grad_norm": 0.169921875,
      "learning_rate": 1.233419858320459e-05,
      "loss": 0.0317,
      "step": 2770
    },
    {
      "epoch": 42.63076923076923,
      "grad_norm": 7.09375,
      "learning_rate": 1.2325472745760686e-05,
      "loss": 0.0289,
      "step": 2771
    },
    {
      "epoch": 42.646153846153844,
      "grad_norm": 1.7890625,
      "learning_rate": 1.2316746993379242e-05,
      "loss": 0.0143,
      "step": 2772
    },
    {
      "epoch": 42.66153846153846,
      "grad_norm": 7.90625,
      "learning_rate": 1.2308021330313089e-05,
      "loss": 0.0287,
      "step": 2773
    },
    {
      "epoch": 42.676923076923075,
      "grad_norm": 0.671875,
      "learning_rate": 1.229929576081501e-05,
      "loss": 0.0101,
      "step": 2774
    },
    {
      "epoch": 42.69230769230769,
      "grad_norm": 2.21875,
      "learning_rate": 1.2290570289137736e-05,
      "loss": 0.0445,
      "step": 2775
    },
    {
      "epoch": 42.707692307692305,
      "grad_norm": 1.9609375,
      "learning_rate": 1.2281844919533957e-05,
      "loss": 0.0254,
      "step": 2776
    },
    {
      "epoch": 42.723076923076924,
      "grad_norm": 1.6875,
      "learning_rate": 1.2273119656256318e-05,
      "loss": 0.019,
      "step": 2777
    },
    {
      "epoch": 42.738461538461536,
      "grad_norm": 2.234375,
      "learning_rate": 1.22643945035574e-05,
      "loss": 0.0181,
      "step": 2778
    },
    {
      "epoch": 42.753846153846155,
      "grad_norm": 2.609375,
      "learning_rate": 1.2255669465689736e-05,
      "loss": 0.0263,
      "step": 2779
    },
    {
      "epoch": 42.76923076923077,
      "grad_norm": 2.046875,
      "learning_rate": 1.2246944546905811e-05,
      "loss": 0.0206,
      "step": 2780
    },
    {
      "epoch": 42.784615384615385,
      "grad_norm": 8.8125,
      "learning_rate": 1.223821975145804e-05,
      "loss": 0.0603,
      "step": 2781
    },
    {
      "epoch": 42.8,
      "grad_norm": 1.734375,
      "learning_rate": 1.2229495083598777e-05,
      "loss": 0.0197,
      "step": 2782
    },
    {
      "epoch": 42.815384615384616,
      "grad_norm": 1.859375,
      "learning_rate": 1.222077054758033e-05,
      "loss": 0.0106,
      "step": 2783
    },
    {
      "epoch": 42.83076923076923,
      "grad_norm": 0.5625,
      "learning_rate": 1.2212046147654926e-05,
      "loss": 0.0119,
      "step": 2784
    },
    {
      "epoch": 42.84615384615385,
      "grad_norm": 2.703125,
      "learning_rate": 1.2203321888074729e-05,
      "loss": 0.0178,
      "step": 2785
    },
    {
      "epoch": 42.86153846153846,
      "grad_norm": 1.5703125,
      "learning_rate": 1.2194597773091837e-05,
      "loss": 0.0112,
      "step": 2786
    },
    {
      "epoch": 42.87692307692308,
      "grad_norm": 2.546875,
      "learning_rate": 1.2185873806958283e-05,
      "loss": 0.0108,
      "step": 2787
    },
    {
      "epoch": 42.89230769230769,
      "grad_norm": 1.5390625,
      "learning_rate": 1.2177149993926017e-05,
      "loss": 0.0205,
      "step": 2788
    },
    {
      "epoch": 42.90769230769231,
      "grad_norm": 0.59765625,
      "learning_rate": 1.2168426338246924e-05,
      "loss": 0.0288,
      "step": 2789
    },
    {
      "epoch": 42.92307692307692,
      "grad_norm": 6.5,
      "learning_rate": 1.2159702844172807e-05,
      "loss": 0.0506,
      "step": 2790
    },
    {
      "epoch": 42.93846153846154,
      "grad_norm": 1.0703125,
      "learning_rate": 1.2150979515955392e-05,
      "loss": 0.0122,
      "step": 2791
    },
    {
      "epoch": 42.95384615384615,
      "grad_norm": 0.65625,
      "learning_rate": 1.2142256357846318e-05,
      "loss": 0.0061,
      "step": 2792
    },
    {
      "epoch": 42.96923076923077,
      "grad_norm": 1.5703125,
      "learning_rate": 1.2133533374097148e-05,
      "loss": 0.0319,
      "step": 2793
    },
    {
      "epoch": 42.98461538461538,
      "grad_norm": 1.515625,
      "learning_rate": 1.2124810568959361e-05,
      "loss": 0.0211,
      "step": 2794
    },
    {
      "epoch": 43.0,
      "grad_norm": 2.875,
      "learning_rate": 1.211608794668434e-05,
      "loss": 0.0176,
      "step": 2795
    },
    {
      "epoch": 43.01538461538462,
      "grad_norm": 3.671875,
      "learning_rate": 1.2107365511523396e-05,
      "loss": 0.0338,
      "step": 2796
    },
    {
      "epoch": 43.03076923076923,
      "grad_norm": 1.2421875,
      "learning_rate": 1.2098643267727732e-05,
      "loss": 0.0069,
      "step": 2797
    },
    {
      "epoch": 43.04615384615385,
      "grad_norm": 2.5,
      "learning_rate": 1.208992121954846e-05,
      "loss": 0.0248,
      "step": 2798
    },
    {
      "epoch": 43.06153846153846,
      "grad_norm": 1.3828125,
      "learning_rate": 1.2081199371236607e-05,
      "loss": 0.0167,
      "step": 2799
    },
    {
      "epoch": 43.07692307692308,
      "grad_norm": 6.875,
      "learning_rate": 1.2072477727043094e-05,
      "loss": 0.0529,
      "step": 2800
    },
    {
      "epoch": 43.09230769230769,
      "grad_norm": 0.75,
      "learning_rate": 1.2063756291218742e-05,
      "loss": 0.0061,
      "step": 2801
    },
    {
      "epoch": 43.10769230769231,
      "grad_norm": 1.0703125,
      "learning_rate": 1.205503506801427e-05,
      "loss": 0.0178,
      "step": 2802
    },
    {
      "epoch": 43.12307692307692,
      "grad_norm": 0.1650390625,
      "learning_rate": 1.2046314061680304e-05,
      "loss": 0.0082,
      "step": 2803
    },
    {
      "epoch": 43.13846153846154,
      "grad_norm": 3.609375,
      "learning_rate": 1.2037593276467353e-05,
      "loss": 0.0415,
      "step": 2804
    },
    {
      "epoch": 43.15384615384615,
      "grad_norm": 2.40625,
      "learning_rate": 1.2028872716625817e-05,
      "loss": 0.0118,
      "step": 2805
    },
    {
      "epoch": 43.16923076923077,
      "grad_norm": 4.625,
      "learning_rate": 1.2020152386406001e-05,
      "loss": 0.0386,
      "step": 2806
    },
    {
      "epoch": 43.184615384615384,
      "grad_norm": 0.4453125,
      "learning_rate": 1.2011432290058082e-05,
      "loss": 0.0104,
      "step": 2807
    },
    {
      "epoch": 43.2,
      "grad_norm": 2.96875,
      "learning_rate": 1.2002712431832127e-05,
      "loss": 0.022,
      "step": 2808
    },
    {
      "epoch": 43.215384615384615,
      "grad_norm": 4.875,
      "learning_rate": 1.1993992815978096e-05,
      "loss": 0.0278,
      "step": 2809
    },
    {
      "epoch": 43.23076923076923,
      "grad_norm": 9.375,
      "learning_rate": 1.1985273446745818e-05,
      "loss": 0.0469,
      "step": 2810
    },
    {
      "epoch": 43.246153846153845,
      "grad_norm": 4.125,
      "learning_rate": 1.1976554328385008e-05,
      "loss": 0.0438,
      "step": 2811
    },
    {
      "epoch": 43.261538461538464,
      "grad_norm": 1.359375,
      "learning_rate": 1.1967835465145256e-05,
      "loss": 0.0235,
      "step": 2812
    },
    {
      "epoch": 43.276923076923076,
      "grad_norm": 1.671875,
      "learning_rate": 1.195911686127604e-05,
      "loss": 0.0128,
      "step": 2813
    },
    {
      "epoch": 43.292307692307695,
      "grad_norm": 4.125,
      "learning_rate": 1.1950398521026691e-05,
      "loss": 0.0352,
      "step": 2814
    },
    {
      "epoch": 43.30769230769231,
      "grad_norm": 1.953125,
      "learning_rate": 1.1941680448646425e-05,
      "loss": 0.0161,
      "step": 2815
    },
    {
      "epoch": 43.323076923076925,
      "grad_norm": 1.0546875,
      "learning_rate": 1.193296264838433e-05,
      "loss": 0.0244,
      "step": 2816
    },
    {
      "epoch": 43.33846153846154,
      "grad_norm": 3.078125,
      "learning_rate": 1.1924245124489345e-05,
      "loss": 0.0234,
      "step": 2817
    },
    {
      "epoch": 43.353846153846156,
      "grad_norm": 1.5234375,
      "learning_rate": 1.1915527881210303e-05,
      "loss": 0.0092,
      "step": 2818
    },
    {
      "epoch": 43.36923076923077,
      "grad_norm": 1.2421875,
      "learning_rate": 1.1906810922795863e-05,
      "loss": 0.02,
      "step": 2819
    },
    {
      "epoch": 43.38461538461539,
      "grad_norm": 3.328125,
      "learning_rate": 1.1898094253494576e-05,
      "loss": 0.0326,
      "step": 2820
    },
    {
      "epoch": 43.4,
      "grad_norm": 5.0,
      "learning_rate": 1.1889377877554835e-05,
      "loss": 0.0335,
      "step": 2821
    },
    {
      "epoch": 43.41538461538462,
      "grad_norm": 2.90625,
      "learning_rate": 1.1880661799224901e-05,
      "loss": 0.0187,
      "step": 2822
    },
    {
      "epoch": 43.43076923076923,
      "grad_norm": 1.6328125,
      "learning_rate": 1.1871946022752882e-05,
      "loss": 0.0096,
      "step": 2823
    },
    {
      "epoch": 43.44615384615385,
      "grad_norm": 0.7578125,
      "learning_rate": 1.186323055238674e-05,
      "loss": 0.0079,
      "step": 2824
    },
    {
      "epoch": 43.46153846153846,
      "grad_norm": 1.703125,
      "learning_rate": 1.1854515392374293e-05,
      "loss": 0.0085,
      "step": 2825
    },
    {
      "epoch": 43.47692307692308,
      "grad_norm": 1.3515625,
      "learning_rate": 1.1845800546963202e-05,
      "loss": 0.0271,
      "step": 2826
    },
    {
      "epoch": 43.49230769230769,
      "grad_norm": 1.703125,
      "learning_rate": 1.183708602040098e-05,
      "loss": 0.0227,
      "step": 2827
    },
    {
      "epoch": 43.50769230769231,
      "grad_norm": 0.87109375,
      "learning_rate": 1.1828371816934974e-05,
      "loss": 0.0077,
      "step": 2828
    },
    {
      "epoch": 43.52307692307692,
      "grad_norm": 1.484375,
      "learning_rate": 1.1819657940812388e-05,
      "loss": 0.0264,
      "step": 2829
    },
    {
      "epoch": 43.53846153846154,
      "grad_norm": 2.484375,
      "learning_rate": 1.1810944396280257e-05,
      "loss": 0.0332,
      "step": 2830
    },
    {
      "epoch": 43.55384615384615,
      "grad_norm": 3.375,
      "learning_rate": 1.1802231187585453e-05,
      "loss": 0.0326,
      "step": 2831
    },
    {
      "epoch": 43.56923076923077,
      "grad_norm": 0.96875,
      "learning_rate": 1.1793518318974694e-05,
      "loss": 0.0059,
      "step": 2832
    },
    {
      "epoch": 43.58461538461538,
      "grad_norm": 2.3125,
      "learning_rate": 1.1784805794694524e-05,
      "loss": 0.0217,
      "step": 2833
    },
    {
      "epoch": 43.6,
      "grad_norm": 1.3828125,
      "learning_rate": 1.177609361899132e-05,
      "loss": 0.0279,
      "step": 2834
    },
    {
      "epoch": 43.61538461538461,
      "grad_norm": 0.341796875,
      "learning_rate": 1.1767381796111293e-05,
      "loss": 0.0121,
      "step": 2835
    },
    {
      "epoch": 43.63076923076923,
      "grad_norm": 4.65625,
      "learning_rate": 1.1758670330300482e-05,
      "loss": 0.0262,
      "step": 2836
    },
    {
      "epoch": 43.646153846153844,
      "grad_norm": 1.015625,
      "learning_rate": 1.1749959225804746e-05,
      "loss": 0.0408,
      "step": 2837
    },
    {
      "epoch": 43.66153846153846,
      "grad_norm": 1.25,
      "learning_rate": 1.1741248486869764e-05,
      "loss": 0.0144,
      "step": 2838
    },
    {
      "epoch": 43.676923076923075,
      "grad_norm": 2.390625,
      "learning_rate": 1.1732538117741058e-05,
      "loss": 0.0329,
      "step": 2839
    },
    {
      "epoch": 43.69230769230769,
      "grad_norm": 3.390625,
      "learning_rate": 1.172382812266395e-05,
      "loss": 0.0429,
      "step": 2840
    },
    {
      "epoch": 43.707692307692305,
      "grad_norm": 1.7265625,
      "learning_rate": 1.1715118505883584e-05,
      "loss": 0.0431,
      "step": 2841
    },
    {
      "epoch": 43.723076923076924,
      "grad_norm": 3.90625,
      "learning_rate": 1.1706409271644925e-05,
      "loss": 0.0179,
      "step": 2842
    },
    {
      "epoch": 43.738461538461536,
      "grad_norm": 3.296875,
      "learning_rate": 1.1697700424192743e-05,
      "loss": 0.0255,
      "step": 2843
    },
    {
      "epoch": 43.753846153846155,
      "grad_norm": 8.3125,
      "learning_rate": 1.1688991967771632e-05,
      "loss": 0.0571,
      "step": 2844
    },
    {
      "epoch": 43.76923076923077,
      "grad_norm": 1.703125,
      "learning_rate": 1.1680283906625985e-05,
      "loss": 0.0253,
      "step": 2845
    },
    {
      "epoch": 43.784615384615385,
      "grad_norm": 2.078125,
      "learning_rate": 1.1671576245000002e-05,
      "loss": 0.0173,
      "step": 2846
    },
    {
      "epoch": 43.8,
      "grad_norm": 3.875,
      "learning_rate": 1.166286898713769e-05,
      "loss": 0.0424,
      "step": 2847
    },
    {
      "epoch": 43.815384615384616,
      "grad_norm": 2.8125,
      "learning_rate": 1.1654162137282867e-05,
      "loss": 0.0654,
      "step": 2848
    },
    {
      "epoch": 43.83076923076923,
      "grad_norm": 1.7109375,
      "learning_rate": 1.1645455699679143e-05,
      "loss": 0.0108,
      "step": 2849
    },
    {
      "epoch": 43.84615384615385,
      "grad_norm": 1.71875,
      "learning_rate": 1.1636749678569928e-05,
      "loss": 0.0095,
      "step": 2850
    },
    {
      "epoch": 43.86153846153846,
      "grad_norm": 1.4453125,
      "learning_rate": 1.1628044078198434e-05,
      "loss": 0.0865,
      "step": 2851
    },
    {
      "epoch": 43.87692307692308,
      "grad_norm": 0.8046875,
      "learning_rate": 1.1619338902807664e-05,
      "loss": 0.0085,
      "step": 2852
    },
    {
      "epoch": 43.89230769230769,
      "grad_norm": 0.7421875,
      "learning_rate": 1.1610634156640419e-05,
      "loss": 0.0147,
      "step": 2853
    },
    {
      "epoch": 43.90769230769231,
      "grad_norm": 1.9609375,
      "learning_rate": 1.1601929843939274e-05,
      "loss": 0.0136,
      "step": 2854
    },
    {
      "epoch": 43.92307692307692,
      "grad_norm": 1.21875,
      "learning_rate": 1.1593225968946617e-05,
      "loss": 0.032,
      "step": 2855
    },
    {
      "epoch": 43.93846153846154,
      "grad_norm": 1.765625,
      "learning_rate": 1.1584522535904607e-05,
      "loss": 0.024,
      "step": 2856
    },
    {
      "epoch": 43.95384615384615,
      "grad_norm": 2.34375,
      "learning_rate": 1.1575819549055188e-05,
      "loss": 0.0153,
      "step": 2857
    },
    {
      "epoch": 43.96923076923077,
      "grad_norm": 1.953125,
      "learning_rate": 1.1567117012640098e-05,
      "loss": 0.0173,
      "step": 2858
    },
    {
      "epoch": 43.98461538461538,
      "grad_norm": 1.1328125,
      "learning_rate": 1.1558414930900843e-05,
      "loss": 0.0096,
      "step": 2859
    },
    {
      "epoch": 44.0,
      "grad_norm": 5.3125,
      "learning_rate": 1.154971330807871e-05,
      "loss": 0.0441,
      "step": 2860
    },
    {
      "epoch": 44.01538461538462,
      "grad_norm": 1.6875,
      "learning_rate": 1.1541012148414768e-05,
      "loss": 0.0103,
      "step": 2861
    },
    {
      "epoch": 44.03076923076923,
      "grad_norm": 1.2578125,
      "learning_rate": 1.1532311456149856e-05,
      "loss": 0.0142,
      "step": 2862
    },
    {
      "epoch": 44.04615384615385,
      "grad_norm": 0.71484375,
      "learning_rate": 1.1523611235524585e-05,
      "loss": 0.0239,
      "step": 2863
    },
    {
      "epoch": 44.06153846153846,
      "grad_norm": 1.4921875,
      "learning_rate": 1.1514911490779333e-05,
      "loss": 0.0122,
      "step": 2864
    },
    {
      "epoch": 44.07692307692308,
      "grad_norm": 0.400390625,
      "learning_rate": 1.1506212226154254e-05,
      "loss": 0.0262,
      "step": 2865
    },
    {
      "epoch": 44.09230769230769,
      "grad_norm": 0.99609375,
      "learning_rate": 1.1497513445889265e-05,
      "loss": 0.014,
      "step": 2866
    },
    {
      "epoch": 44.10769230769231,
      "grad_norm": 3.390625,
      "learning_rate": 1.1488815154224042e-05,
      "loss": 0.0203,
      "step": 2867
    },
    {
      "epoch": 44.12307692307692,
      "grad_norm": 1.6015625,
      "learning_rate": 1.1480117355398032e-05,
      "loss": 0.0083,
      "step": 2868
    },
    {
      "epoch": 44.13846153846154,
      "grad_norm": 0.7734375,
      "learning_rate": 1.147142005365043e-05,
      "loss": 0.0193,
      "step": 2869
    },
    {
      "epoch": 44.15384615384615,
      "grad_norm": 3.8125,
      "learning_rate": 1.14627232532202e-05,
      "loss": 0.0338,
      "step": 2870
    },
    {
      "epoch": 44.16923076923077,
      "grad_norm": 0.365234375,
      "learning_rate": 1.1454026958346058e-05,
      "loss": 0.0195,
      "step": 2871
    },
    {
      "epoch": 44.184615384615384,
      "grad_norm": 0.86328125,
      "learning_rate": 1.1445331173266468e-05,
      "loss": 0.0062,
      "step": 2872
    },
    {
      "epoch": 44.2,
      "grad_norm": 2.0,
      "learning_rate": 1.143663590221965e-05,
      "loss": 0.0209,
      "step": 2873
    },
    {
      "epoch": 44.215384615384615,
      "grad_norm": 2.59375,
      "learning_rate": 1.1427941149443578e-05,
      "loss": 0.0241,
      "step": 2874
    },
    {
      "epoch": 44.23076923076923,
      "grad_norm": 2.03125,
      "learning_rate": 1.1419246919175966e-05,
      "loss": 0.0123,
      "step": 2875
    },
    {
      "epoch": 44.246153846153845,
      "grad_norm": 3.25,
      "learning_rate": 1.1410553215654272e-05,
      "loss": 0.0207,
      "step": 2876
    },
    {
      "epoch": 44.261538461538464,
      "grad_norm": 2.6875,
      "learning_rate": 1.140186004311571e-05,
      "loss": 0.0261,
      "step": 2877
    },
    {
      "epoch": 44.276923076923076,
      "grad_norm": 1.515625,
      "learning_rate": 1.1393167405797223e-05,
      "loss": 0.0105,
      "step": 2878
    },
    {
      "epoch": 44.292307692307695,
      "grad_norm": 1.0625,
      "learning_rate": 1.1384475307935493e-05,
      "loss": 0.0173,
      "step": 2879
    },
    {
      "epoch": 44.30769230769231,
      "grad_norm": 1.6015625,
      "learning_rate": 1.1375783753766953e-05,
      "loss": 0.0453,
      "step": 2880
    },
    {
      "epoch": 44.323076923076925,
      "grad_norm": 1.3046875,
      "learning_rate": 1.1367092747527752e-05,
      "loss": 0.0087,
      "step": 2881
    },
    {
      "epoch": 44.33846153846154,
      "grad_norm": 5.65625,
      "learning_rate": 1.135840229345378e-05,
      "loss": 0.0315,
      "step": 2882
    },
    {
      "epoch": 44.353846153846156,
      "grad_norm": 0.60546875,
      "learning_rate": 1.1349712395780663e-05,
      "loss": 0.0124,
      "step": 2883
    },
    {
      "epoch": 44.36923076923077,
      "grad_norm": 0.77734375,
      "learning_rate": 1.1341023058743749e-05,
      "loss": 0.0064,
      "step": 2884
    },
    {
      "epoch": 44.38461538461539,
      "grad_norm": 0.87109375,
      "learning_rate": 1.133233428657812e-05,
      "loss": 0.0517,
      "step": 2885
    },
    {
      "epoch": 44.4,
      "grad_norm": 2.828125,
      "learning_rate": 1.1323646083518572e-05,
      "loss": 0.0219,
      "step": 2886
    },
    {
      "epoch": 44.41538461538462,
      "grad_norm": 0.8046875,
      "learning_rate": 1.1314958453799634e-05,
      "loss": 0.0052,
      "step": 2887
    },
    {
      "epoch": 44.43076923076923,
      "grad_norm": 4.625,
      "learning_rate": 1.1306271401655547e-05,
      "loss": 0.0444,
      "step": 2888
    },
    {
      "epoch": 44.44615384615385,
      "grad_norm": 3.484375,
      "learning_rate": 1.1297584931320284e-05,
      "loss": 0.0215,
      "step": 2889
    },
    {
      "epoch": 44.46153846153846,
      "grad_norm": 4.34375,
      "learning_rate": 1.1288899047027512e-05,
      "loss": 0.0306,
      "step": 2890
    },
    {
      "epoch": 44.47692307692308,
      "grad_norm": 2.25,
      "learning_rate": 1.1280213753010633e-05,
      "loss": 0.0092,
      "step": 2891
    },
    {
      "epoch": 44.49230769230769,
      "grad_norm": 3.75,
      "learning_rate": 1.1271529053502755e-05,
      "loss": 0.0223,
      "step": 2892
    },
    {
      "epoch": 44.50769230769231,
      "grad_norm": 3.359375,
      "learning_rate": 1.1262844952736688e-05,
      "loss": 0.0467,
      "step": 2893
    },
    {
      "epoch": 44.52307692307692,
      "grad_norm": 3.21875,
      "learning_rate": 1.1254161454944965e-05,
      "loss": 0.0366,
      "step": 2894
    },
    {
      "epoch": 44.53846153846154,
      "grad_norm": 0.72265625,
      "learning_rate": 1.1245478564359812e-05,
      "loss": 0.0112,
      "step": 2895
    },
    {
      "epoch": 44.55384615384615,
      "grad_norm": 1.84375,
      "learning_rate": 1.123679628521317e-05,
      "loss": 0.0361,
      "step": 2896
    },
    {
      "epoch": 44.56923076923077,
      "grad_norm": 5.1875,
      "learning_rate": 1.1228114621736673e-05,
      "loss": 0.0353,
      "step": 2897
    },
    {
      "epoch": 44.58461538461538,
      "grad_norm": 0.3828125,
      "learning_rate": 1.121943357816166e-05,
      "loss": 0.0046,
      "step": 2898
    },
    {
      "epoch": 44.6,
      "grad_norm": 0.65625,
      "learning_rate": 1.1210753158719162e-05,
      "loss": 0.0043,
      "step": 2899
    },
    {
      "epoch": 44.61538461538461,
      "grad_norm": 4.1875,
      "learning_rate": 1.1202073367639911e-05,
      "loss": 0.0289,
      "step": 2900
    },
    {
      "epoch": 44.63076923076923,
      "grad_norm": 1.75,
      "learning_rate": 1.1193394209154334e-05,
      "loss": 0.017,
      "step": 2901
    },
    {
      "epoch": 44.646153846153844,
      "grad_norm": 1.1328125,
      "learning_rate": 1.1184715687492544e-05,
      "loss": 0.0497,
      "step": 2902
    },
    {
      "epoch": 44.66153846153846,
      "grad_norm": 0.92578125,
      "learning_rate": 1.1176037806884354e-05,
      "loss": 0.0112,
      "step": 2903
    },
    {
      "epoch": 44.676923076923075,
      "grad_norm": 2.296875,
      "learning_rate": 1.1167360571559252e-05,
      "loss": 0.0264,
      "step": 2904
    },
    {
      "epoch": 44.69230769230769,
      "grad_norm": 3.265625,
      "learning_rate": 1.1158683985746418e-05,
      "loss": 0.0255,
      "step": 2905
    },
    {
      "epoch": 44.707692307692305,
      "grad_norm": 1.90625,
      "learning_rate": 1.1150008053674715e-05,
      "loss": 0.0161,
      "step": 2906
    },
    {
      "epoch": 44.723076923076924,
      "grad_norm": 1.3125,
      "learning_rate": 1.1141332779572689e-05,
      "loss": 0.0455,
      "step": 2907
    },
    {
      "epoch": 44.738461538461536,
      "grad_norm": 0.181640625,
      "learning_rate": 1.113265816766856e-05,
      "loss": 0.0077,
      "step": 2908
    },
    {
      "epoch": 44.753846153846155,
      "grad_norm": 1.6875,
      "learning_rate": 1.1123984222190225e-05,
      "loss": 0.0094,
      "step": 2909
    },
    {
      "epoch": 44.76923076923077,
      "grad_norm": 2.125,
      "learning_rate": 1.1115310947365268e-05,
      "loss": 0.0368,
      "step": 2910
    },
    {
      "epoch": 44.784615384615385,
      "grad_norm": 2.90625,
      "learning_rate": 1.1106638347420933e-05,
      "loss": 0.0228,
      "step": 2911
    },
    {
      "epoch": 44.8,
      "grad_norm": 0.6328125,
      "learning_rate": 1.1097966426584139e-05,
      "loss": 0.0079,
      "step": 2912
    },
    {
      "epoch": 44.815384615384616,
      "grad_norm": 4.09375,
      "learning_rate": 1.108929518908148e-05,
      "loss": 0.0232,
      "step": 2913
    },
    {
      "epoch": 44.83076923076923,
      "grad_norm": 0.66015625,
      "learning_rate": 1.1080624639139203e-05,
      "loss": 0.0088,
      "step": 2914
    },
    {
      "epoch": 44.84615384615385,
      "grad_norm": 6.5,
      "learning_rate": 1.1071954780983245e-05,
      "loss": 0.0384,
      "step": 2915
    },
    {
      "epoch": 44.86153846153846,
      "grad_norm": 0.48046875,
      "learning_rate": 1.106328561883917e-05,
      "loss": 0.0029,
      "step": 2916
    },
    {
      "epoch": 44.87692307692308,
      "grad_norm": 0.8359375,
      "learning_rate": 1.1054617156932233e-05,
      "loss": 0.0104,
      "step": 2917
    },
    {
      "epoch": 44.89230769230769,
      "grad_norm": 2.5625,
      "learning_rate": 1.1045949399487334e-05,
      "loss": 0.0257,
      "step": 2918
    },
    {
      "epoch": 44.90769230769231,
      "grad_norm": 2.609375,
      "learning_rate": 1.1037282350729033e-05,
      "loss": 0.0238,
      "step": 2919
    },
    {
      "epoch": 44.92307692307692,
      "grad_norm": 4.6875,
      "learning_rate": 1.1028616014881546e-05,
      "loss": 0.0321,
      "step": 2920
    },
    {
      "epoch": 44.93846153846154,
      "grad_norm": 1.7578125,
      "learning_rate": 1.1019950396168737e-05,
      "loss": 0.0197,
      "step": 2921
    },
    {
      "epoch": 44.95384615384615,
      "grad_norm": 0.890625,
      "learning_rate": 1.1011285498814128e-05,
      "loss": 0.0163,
      "step": 2922
    },
    {
      "epoch": 44.96923076923077,
      "grad_norm": 3.28125,
      "learning_rate": 1.100262132704088e-05,
      "loss": 0.0284,
      "step": 2923
    },
    {
      "epoch": 44.98461538461538,
      "grad_norm": 1.3046875,
      "learning_rate": 1.0993957885071812e-05,
      "loss": 0.0065,
      "step": 2924
    },
    {
      "epoch": 45.0,
      "grad_norm": 4.21875,
      "learning_rate": 1.0985295177129373e-05,
      "loss": 0.0198,
      "step": 2925
    },
    {
      "epoch": 45.01538461538462,
      "grad_norm": 0.69140625,
      "learning_rate": 1.097663320743566e-05,
      "loss": 0.0076,
      "step": 2926
    },
    {
      "epoch": 45.03076923076923,
      "grad_norm": 4.46875,
      "learning_rate": 1.0967971980212417e-05,
      "loss": 0.0829,
      "step": 2927
    },
    {
      "epoch": 45.04615384615385,
      "grad_norm": 0.5703125,
      "learning_rate": 1.0959311499681018e-05,
      "loss": 0.0168,
      "step": 2928
    },
    {
      "epoch": 45.06153846153846,
      "grad_norm": 6.03125,
      "learning_rate": 1.0950651770062478e-05,
      "loss": 0.0219,
      "step": 2929
    },
    {
      "epoch": 45.07692307692308,
      "grad_norm": 1.7734375,
      "learning_rate": 1.0941992795577447e-05,
      "loss": 0.0154,
      "step": 2930
    },
    {
      "epoch": 45.09230769230769,
      "grad_norm": 3.984375,
      "learning_rate": 1.0933334580446195e-05,
      "loss": 0.0323,
      "step": 2931
    },
    {
      "epoch": 45.10769230769231,
      "grad_norm": 9.5,
      "learning_rate": 1.092467712888864e-05,
      "loss": 0.0451,
      "step": 2932
    },
    {
      "epoch": 45.12307692307692,
      "grad_norm": 0.41015625,
      "learning_rate": 1.091602044512432e-05,
      "loss": 0.008,
      "step": 2933
    },
    {
      "epoch": 45.13846153846154,
      "grad_norm": 2.1875,
      "learning_rate": 1.0907364533372388e-05,
      "loss": 0.0147,
      "step": 2934
    },
    {
      "epoch": 45.15384615384615,
      "grad_norm": 1.171875,
      "learning_rate": 1.0898709397851634e-05,
      "loss": 0.0142,
      "step": 2935
    },
    {
      "epoch": 45.16923076923077,
      "grad_norm": 2.203125,
      "learning_rate": 1.0890055042780467e-05,
      "loss": 0.0414,
      "step": 2936
    },
    {
      "epoch": 45.184615384615384,
      "grad_norm": 2.03125,
      "learning_rate": 1.0881401472376915e-05,
      "loss": 0.0241,
      "step": 2937
    },
    {
      "epoch": 45.2,
      "grad_norm": 2.640625,
      "learning_rate": 1.087274869085862e-05,
      "loss": 0.0483,
      "step": 2938
    },
    {
      "epoch": 45.215384615384615,
      "grad_norm": 3.21875,
      "learning_rate": 1.0864096702442848e-05,
      "loss": 0.0285,
      "step": 2939
    },
    {
      "epoch": 45.23076923076923,
      "grad_norm": 0.86328125,
      "learning_rate": 1.085544551134647e-05,
      "loss": 0.0085,
      "step": 2940
    },
    {
      "epoch": 45.246153846153845,
      "grad_norm": 4.21875,
      "learning_rate": 1.0846795121785969e-05,
      "loss": 0.028,
      "step": 2941
    },
    {
      "epoch": 45.261538461538464,
      "grad_norm": 1.609375,
      "learning_rate": 1.0838145537977447e-05,
      "loss": 0.0206,
      "step": 2942
    },
    {
      "epoch": 45.276923076923076,
      "grad_norm": 1.359375,
      "learning_rate": 1.0829496764136598e-05,
      "loss": 0.0107,
      "step": 2943
    },
    {
      "epoch": 45.292307692307695,
      "grad_norm": 3.28125,
      "learning_rate": 1.0820848804478733e-05,
      "loss": 0.0187,
      "step": 2944
    },
    {
      "epoch": 45.30769230769231,
      "grad_norm": 8.5,
      "learning_rate": 1.0812201663218762e-05,
      "loss": 0.0442,
      "step": 2945
    },
    {
      "epoch": 45.323076923076925,
      "grad_norm": 2.421875,
      "learning_rate": 1.0803555344571197e-05,
      "loss": 0.0127,
      "step": 2946
    },
    {
      "epoch": 45.33846153846154,
      "grad_norm": 1.3515625,
      "learning_rate": 1.0794909852750145e-05,
      "loss": 0.019,
      "step": 2947
    },
    {
      "epoch": 45.353846153846156,
      "grad_norm": 1.15625,
      "learning_rate": 1.0786265191969322e-05,
      "loss": 0.0224,
      "step": 2948
    },
    {
      "epoch": 45.36923076923077,
      "grad_norm": 0.376953125,
      "learning_rate": 1.0777621366442026e-05,
      "loss": 0.0122,
      "step": 2949
    },
    {
      "epoch": 45.38461538461539,
      "grad_norm": 0.8125,
      "learning_rate": 1.0768978380381148e-05,
      "loss": 0.0118,
      "step": 2950
    },
    {
      "epoch": 45.4,
      "grad_norm": 0.5,
      "learning_rate": 1.0760336237999187e-05,
      "loss": 0.0449,
      "step": 2951
    },
    {
      "epoch": 45.41538461538462,
      "grad_norm": 3.0625,
      "learning_rate": 1.0751694943508202e-05,
      "loss": 0.026,
      "step": 2952
    },
    {
      "epoch": 45.43076923076923,
      "grad_norm": 7.0,
      "learning_rate": 1.0743054501119865e-05,
      "loss": 0.0202,
      "step": 2953
    },
    {
      "epoch": 45.44615384615385,
      "grad_norm": 1.0078125,
      "learning_rate": 1.0734414915045419e-05,
      "loss": 0.0543,
      "step": 2954
    },
    {
      "epoch": 45.46153846153846,
      "grad_norm": 4.34375,
      "learning_rate": 1.0725776189495696e-05,
      "loss": 0.0394,
      "step": 2955
    },
    {
      "epoch": 45.47692307692308,
      "grad_norm": 0.33984375,
      "learning_rate": 1.0717138328681106e-05,
      "loss": 0.0047,
      "step": 2956
    },
    {
      "epoch": 45.49230769230769,
      "grad_norm": 5.65625,
      "learning_rate": 1.0708501336811632e-05,
      "loss": 0.0402,
      "step": 2957
    },
    {
      "epoch": 45.50769230769231,
      "grad_norm": 4.125,
      "learning_rate": 1.0699865218096849e-05,
      "loss": 0.0391,
      "step": 2958
    },
    {
      "epoch": 45.52307692307692,
      "grad_norm": 3.890625,
      "learning_rate": 1.0691229976745887e-05,
      "loss": 0.0388,
      "step": 2959
    },
    {
      "epoch": 45.53846153846154,
      "grad_norm": 2.828125,
      "learning_rate": 1.0682595616967465e-05,
      "loss": 0.0202,
      "step": 2960
    },
    {
      "epoch": 45.55384615384615,
      "grad_norm": 8.25,
      "learning_rate": 1.0673962142969857e-05,
      "loss": 0.0863,
      "step": 2961
    },
    {
      "epoch": 45.56923076923077,
      "grad_norm": 1.25,
      "learning_rate": 1.0665329558960918e-05,
      "loss": 0.0421,
      "step": 2962
    },
    {
      "epoch": 45.58461538461538,
      "grad_norm": 1.65625,
      "learning_rate": 1.0656697869148067e-05,
      "loss": 0.0075,
      "step": 2963
    },
    {
      "epoch": 45.6,
      "grad_norm": 0.6953125,
      "learning_rate": 1.0648067077738275e-05,
      "loss": 0.0096,
      "step": 2964
    },
    {
      "epoch": 45.61538461538461,
      "grad_norm": 1.765625,
      "learning_rate": 1.0639437188938098e-05,
      "loss": 0.0196,
      "step": 2965
    },
    {
      "epoch": 45.63076923076923,
      "grad_norm": 7.71875,
      "learning_rate": 1.0630808206953632e-05,
      "loss": 0.0462,
      "step": 2966
    },
    {
      "epoch": 45.646153846153844,
      "grad_norm": 2.65625,
      "learning_rate": 1.0622180135990537e-05,
      "loss": 0.0222,
      "step": 2967
    },
    {
      "epoch": 45.66153846153846,
      "grad_norm": 1.7734375,
      "learning_rate": 1.0613552980254032e-05,
      "loss": 0.0128,
      "step": 2968
    },
    {
      "epoch": 45.676923076923075,
      "grad_norm": 2.65625,
      "learning_rate": 1.0604926743948892e-05,
      "loss": 0.0095,
      "step": 2969
    },
    {
      "epoch": 45.69230769230769,
      "grad_norm": 0.89453125,
      "learning_rate": 1.0596301431279432e-05,
      "loss": 0.0206,
      "step": 2970
    },
    {
      "epoch": 45.707692307692305,
      "grad_norm": 1.75,
      "learning_rate": 1.0587677046449526e-05,
      "loss": 0.0172,
      "step": 2971
    },
    {
      "epoch": 45.723076923076924,
      "grad_norm": 1.125,
      "learning_rate": 1.0579053593662599e-05,
      "loss": 0.0135,
      "step": 2972
    },
    {
      "epoch": 45.738461538461536,
      "grad_norm": 4.25,
      "learning_rate": 1.057043107712161e-05,
      "loss": 0.0246,
      "step": 2973
    },
    {
      "epoch": 45.753846153846155,
      "grad_norm": 3.03125,
      "learning_rate": 1.0561809501029078e-05,
      "loss": 0.0167,
      "step": 2974
    },
    {
      "epoch": 45.76923076923077,
      "grad_norm": 4.28125,
      "learning_rate": 1.0553188869587049e-05,
      "loss": 0.0373,
      "step": 2975
    },
    {
      "epoch": 45.784615384615385,
      "grad_norm": 0.37109375,
      "learning_rate": 1.0544569186997112e-05,
      "loss": 0.0047,
      "step": 2976
    },
    {
      "epoch": 45.8,
      "grad_norm": 1.7578125,
      "learning_rate": 1.0535950457460408e-05,
      "loss": 0.0178,
      "step": 2977
    },
    {
      "epoch": 45.815384615384616,
      "grad_norm": 1.2265625,
      "learning_rate": 1.0527332685177583e-05,
      "loss": 0.0279,
      "step": 2978
    },
    {
      "epoch": 45.83076923076923,
      "grad_norm": 0.55859375,
      "learning_rate": 1.0518715874348844e-05,
      "loss": 0.0134,
      "step": 2979
    },
    {
      "epoch": 45.84615384615385,
      "grad_norm": 1.125,
      "learning_rate": 1.0510100029173917e-05,
      "loss": 0.0122,
      "step": 2980
    },
    {
      "epoch": 45.86153846153846,
      "grad_norm": 3.65625,
      "learning_rate": 1.0501485153852066e-05,
      "loss": 0.0158,
      "step": 2981
    },
    {
      "epoch": 45.87692307692308,
      "grad_norm": 3.0,
      "learning_rate": 1.0492871252582072e-05,
      "loss": 0.0189,
      "step": 2982
    },
    {
      "epoch": 45.89230769230769,
      "grad_norm": 3.859375,
      "learning_rate": 1.0484258329562243e-05,
      "loss": 0.0169,
      "step": 2983
    },
    {
      "epoch": 45.90769230769231,
      "grad_norm": 0.189453125,
      "learning_rate": 1.047564638899042e-05,
      "loss": 0.0163,
      "step": 2984
    },
    {
      "epoch": 45.92307692307692,
      "grad_norm": 1.3984375,
      "learning_rate": 1.0467035435063956e-05,
      "loss": 0.0082,
      "step": 2985
    },
    {
      "epoch": 45.93846153846154,
      "grad_norm": 1.59375,
      "learning_rate": 1.0458425471979726e-05,
      "loss": 0.0091,
      "step": 2986
    },
    {
      "epoch": 45.95384615384615,
      "grad_norm": 2.171875,
      "learning_rate": 1.0449816503934114e-05,
      "loss": 0.0152,
      "step": 2987
    },
    {
      "epoch": 45.96923076923077,
      "grad_norm": 4.34375,
      "learning_rate": 1.0441208535123035e-05,
      "loss": 0.0369,
      "step": 2988
    },
    {
      "epoch": 45.98461538461538,
      "grad_norm": 0.609375,
      "learning_rate": 1.0432601569741906e-05,
      "loss": 0.0355,
      "step": 2989
    },
    {
      "epoch": 46.0,
      "grad_norm": 0.234375,
      "learning_rate": 1.0423995611985653e-05,
      "loss": 0.004,
      "step": 2990
    },
    {
      "epoch": 46.01538461538462,
      "grad_norm": 3.78125,
      "learning_rate": 1.0415390666048723e-05,
      "loss": 0.0368,
      "step": 2991
    },
    {
      "epoch": 46.03076923076923,
      "grad_norm": 2.671875,
      "learning_rate": 1.0406786736125057e-05,
      "loss": 0.0308,
      "step": 2992
    },
    {
      "epoch": 46.04615384615385,
      "grad_norm": 3.625,
      "learning_rate": 1.0398183826408107e-05,
      "loss": 0.0359,
      "step": 2993
    },
    {
      "epoch": 46.06153846153846,
      "grad_norm": 1.03125,
      "learning_rate": 1.0389581941090827e-05,
      "loss": 0.0077,
      "step": 2994
    },
    {
      "epoch": 46.07692307692308,
      "grad_norm": 2.5,
      "learning_rate": 1.0380981084365675e-05,
      "loss": 0.0112,
      "step": 2995
    },
    {
      "epoch": 46.09230769230769,
      "grad_norm": 1.6796875,
      "learning_rate": 1.0372381260424597e-05,
      "loss": 0.0093,
      "step": 2996
    },
    {
      "epoch": 46.10769230769231,
      "grad_norm": 1.9140625,
      "learning_rate": 1.0363782473459044e-05,
      "loss": 0.0159,
      "step": 2997
    },
    {
      "epoch": 46.12307692307692,
      "grad_norm": 0.326171875,
      "learning_rate": 1.0355184727659964e-05,
      "loss": 0.0064,
      "step": 2998
    },
    {
      "epoch": 46.13846153846154,
      "grad_norm": 0.1552734375,
      "learning_rate": 1.0346588027217793e-05,
      "loss": 0.0215,
      "step": 2999
    },
    {
      "epoch": 46.15384615384615,
      "grad_norm": 2.765625,
      "learning_rate": 1.0337992376322456e-05,
      "loss": 0.0199,
      "step": 3000
    }
  ],
  "logging_steps": 1,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 77,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}