{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 155,
  "global_step": 620,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 5.8125,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.7582,
      "step": 1
    },
    {
      "epoch": 0.0,
      "eval_loss": 2.128222703933716,
      "eval_runtime": 1.7853,
      "eval_samples_per_second": 19.044,
      "eval_steps_per_second": 19.044,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.0625,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.0767,
      "step": 2
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.875,
      "learning_rate": 6e-06,
      "loss": 2.0746,
      "step": 3
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.1875,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.9598,
      "step": 4
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.75,
      "learning_rate": 1e-05,
      "loss": 2.1597,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.15625,
      "learning_rate": 1.2e-05,
      "loss": 1.5354,
      "step": 6
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.53125,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 1.7962,
      "step": 7
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.875,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.0638,
      "step": 8
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.4375,
      "learning_rate": 1.8e-05,
      "loss": 1.5363,
      "step": 9
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.125,
      "learning_rate": 2e-05,
      "loss": 2.2466,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.9375,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 1.8931,
      "step": 11
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.0625,
      "learning_rate": 2.4e-05,
      "loss": 1.579,
      "step": 12
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.9375,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 1.5665,
      "step": 13
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.25,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 1.9216,
      "step": 14
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.78125,
      "learning_rate": 3e-05,
      "loss": 2.0173,
      "step": 15
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.5,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 2.0586,
      "step": 16
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.84375,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 1.9965,
      "step": 17
    },
    {
      "epoch": 0.03,
      "grad_norm": 28.125,
      "learning_rate": 3.6e-05,
      "loss": 2.2671,
      "step": 18
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.5,
      "learning_rate": 3.8e-05,
      "loss": 2.3432,
      "step": 19
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.1875,
      "learning_rate": 4e-05,
      "loss": 2.3057,
      "step": 20
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.6875,
      "learning_rate": 4.2e-05,
      "loss": 2.1117,
      "step": 21
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.25,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 2.8508,
      "step": 22
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.0,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.9681,
      "step": 23
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.59375,
      "learning_rate": 4.8e-05,
      "loss": 2.3675,
      "step": 24
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.40625,
      "learning_rate": 5e-05,
      "loss": 1.8022,
      "step": 25
    },
    {
      "epoch": 0.04,
      "grad_norm": 17.125,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 2.0862,
      "step": 26
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.40625,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 1.8344,
      "step": 27
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.40625,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 2.0822,
      "step": 28
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.25,
      "learning_rate": 5.8e-05,
      "loss": 2.3784,
      "step": 29
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.875,
      "learning_rate": 6e-05,
      "loss": 1.8403,
      "step": 30
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.8125,
      "learning_rate": 6.2e-05,
      "loss": 1.7077,
      "step": 31
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.65625,
      "learning_rate": 6.400000000000001e-05,
      "loss": 1.8347,
      "step": 32
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.15625,
      "learning_rate": 6.6e-05,
      "loss": 1.9535,
      "step": 33
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.34375,
      "learning_rate": 6.800000000000001e-05,
      "loss": 2.3192,
      "step": 34
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.46875,
      "learning_rate": 7e-05,
      "loss": 2.2975,
      "step": 35
    },
    {
      "epoch": 0.06,
      "grad_norm": 6.6875,
      "learning_rate": 7.2e-05,
      "loss": 2.2092,
      "step": 36
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.4375,
      "learning_rate": 7.4e-05,
      "loss": 1.872,
      "step": 37
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.0625,
      "learning_rate": 7.6e-05,
      "loss": 2.8407,
      "step": 38
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.8125,
      "learning_rate": 7.800000000000001e-05,
      "loss": 2.1849,
      "step": 39
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.375,
      "learning_rate": 8e-05,
      "loss": 1.7414,
      "step": 40
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.375,
      "learning_rate": 8.2e-05,
      "loss": 2.3472,
      "step": 41
    },
    {
      "epoch": 0.07,
      "grad_norm": 5.9375,
      "learning_rate": 8.4e-05,
      "loss": 1.6975,
      "step": 42
    },
    {
      "epoch": 0.07,
      "grad_norm": 5.78125,
      "learning_rate": 8.6e-05,
      "loss": 1.9298,
      "step": 43
    },
    {
      "epoch": 0.07,
      "grad_norm": 6.34375,
      "learning_rate": 8.800000000000001e-05,
      "loss": 1.8339,
      "step": 44
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.0,
      "learning_rate": 9e-05,
      "loss": 1.4134,
      "step": 45
    },
    {
      "epoch": 0.07,
      "grad_norm": 6.0,
      "learning_rate": 9.200000000000001e-05,
      "loss": 1.656,
      "step": 46
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.0,
      "learning_rate": 9.4e-05,
      "loss": 2.4637,
      "step": 47
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.8125,
      "learning_rate": 9.6e-05,
      "loss": 1.8719,
      "step": 48
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.5625,
      "learning_rate": 9.8e-05,
      "loss": 2.3955,
      "step": 49
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.0625,
      "learning_rate": 0.0001,
      "loss": 2.024,
      "step": 50
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.0625,
      "learning_rate": 0.00010200000000000001,
      "loss": 2.2503,
      "step": 51
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.4375,
      "learning_rate": 0.00010400000000000001,
      "loss": 2.2305,
      "step": 52
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.6875,
      "learning_rate": 0.00010600000000000002,
      "loss": 2.0418,
      "step": 53
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.25,
      "learning_rate": 0.00010800000000000001,
      "loss": 2.4381,
      "step": 54
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.53125,
      "learning_rate": 0.00011000000000000002,
      "loss": 1.9802,
      "step": 55
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.75,
      "learning_rate": 0.00011200000000000001,
      "loss": 1.9867,
      "step": 56
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.21875,
      "learning_rate": 0.00011399999999999999,
      "loss": 2.0959,
      "step": 57
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.34375,
      "learning_rate": 0.000116,
      "loss": 2.0148,
      "step": 58
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.75,
      "learning_rate": 0.000118,
      "loss": 1.8376,
      "step": 59
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.59375,
      "learning_rate": 0.00012,
      "loss": 2.0351,
      "step": 60
    },
    {
      "epoch": 0.1,
      "grad_norm": 147.0,
      "learning_rate": 0.000122,
      "loss": 7.0141,
      "step": 61
    },
    {
      "epoch": 0.1,
      "grad_norm": 15.3125,
      "learning_rate": 0.000124,
      "loss": 2.4206,
      "step": 62
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.9375,
      "learning_rate": 0.000126,
      "loss": 2.049,
      "step": 63
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.875,
      "learning_rate": 0.00012800000000000002,
      "loss": 2.7425,
      "step": 64
    },
    {
      "epoch": 0.1,
      "grad_norm": 6.90625,
      "learning_rate": 0.00013000000000000002,
      "loss": 1.6744,
      "step": 65
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.25,
      "learning_rate": 0.000132,
      "loss": 2.9067,
      "step": 66
    },
    {
      "epoch": 0.11,
      "grad_norm": 6.78125,
      "learning_rate": 0.000134,
      "loss": 2.0443,
      "step": 67
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.8125,
      "learning_rate": 0.00013600000000000003,
      "loss": 2.1493,
      "step": 68
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.4375,
      "learning_rate": 0.000138,
      "loss": 2.4634,
      "step": 69
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.125,
      "learning_rate": 0.00014,
      "loss": 2.5528,
      "step": 70
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.625,
      "learning_rate": 0.000142,
      "loss": 2.8141,
      "step": 71
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.625,
      "learning_rate": 0.000144,
      "loss": 2.3318,
      "step": 72
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.78125,
      "learning_rate": 0.000146,
      "loss": 2.2342,
      "step": 73
    },
    {
      "epoch": 0.12,
      "grad_norm": 440.0,
      "learning_rate": 0.000148,
      "loss": 11.3354,
      "step": 74
    },
    {
      "epoch": 0.12,
      "grad_norm": 16.75,
      "learning_rate": 0.00015000000000000001,
      "loss": 2.4637,
      "step": 75
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.3125,
      "learning_rate": 0.000152,
      "loss": 3.3555,
      "step": 76
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.5625,
      "learning_rate": 0.000154,
      "loss": 2.9384,
      "step": 77
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.5625,
      "learning_rate": 0.00015600000000000002,
      "loss": 2.7772,
      "step": 78
    },
    {
      "epoch": 0.13,
      "grad_norm": 6.375,
      "learning_rate": 0.00015800000000000002,
      "loss": 2.2272,
      "step": 79
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.9375,
      "learning_rate": 0.00016,
      "loss": 2.5675,
      "step": 80
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.0,
      "learning_rate": 0.000162,
      "loss": 3.1618,
      "step": 81
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.9375,
      "learning_rate": 0.000164,
      "loss": 2.2929,
      "step": 82
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.125,
      "learning_rate": 0.000166,
      "loss": 2.1003,
      "step": 83
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.6875,
      "learning_rate": 0.000168,
      "loss": 2.3771,
      "step": 84
    },
    {
      "epoch": 0.14,
      "grad_norm": 12.25,
      "learning_rate": 0.00017,
      "loss": 2.8536,
      "step": 85
    },
    {
      "epoch": 0.14,
      "grad_norm": 6.0625,
      "learning_rate": 0.000172,
      "loss": 2.3478,
      "step": 86
    },
    {
      "epoch": 0.14,
      "grad_norm": 23.875,
      "learning_rate": 0.000174,
      "loss": 3.0549,
      "step": 87
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.25,
      "learning_rate": 0.00017600000000000002,
      "loss": 2.5753,
      "step": 88
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.5,
      "learning_rate": 0.00017800000000000002,
      "loss": 2.2508,
      "step": 89
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.875,
      "learning_rate": 0.00018,
      "loss": 2.476,
      "step": 90
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.3125,
      "learning_rate": 0.000182,
      "loss": 2.7688,
      "step": 91
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.0625,
      "learning_rate": 0.00018400000000000003,
      "loss": 2.7879,
      "step": 92
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.25,
      "learning_rate": 0.00018600000000000002,
      "loss": 2.1463,
      "step": 93
    },
    {
      "epoch": 0.15,
      "grad_norm": 16.5,
      "learning_rate": 0.000188,
      "loss": 2.5949,
      "step": 94
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.8125,
      "learning_rate": 0.00019,
      "loss": 2.381,
      "step": 95
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.0625,
      "learning_rate": 0.000192,
      "loss": 2.592,
      "step": 96
    },
    {
      "epoch": 0.16,
      "grad_norm": 16.375,
      "learning_rate": 0.000194,
      "loss": 2.6753,
      "step": 97
    },
    {
      "epoch": 0.16,
      "grad_norm": 55.0,
      "learning_rate": 0.000196,
      "loss": 2.5502,
      "step": 98
    },
    {
      "epoch": 0.16,
      "grad_norm": 24.5,
      "learning_rate": 0.00019800000000000002,
      "loss": 2.4627,
      "step": 99
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.875,
      "learning_rate": 0.0002,
      "loss": 2.0743,
      "step": 100
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.34375,
      "learning_rate": 0.00019999817500473724,
      "loss": 2.3793,
      "step": 101
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.125,
      "learning_rate": 0.00019999270008556108,
      "loss": 2.6009,
      "step": 102
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.125,
      "learning_rate": 0.00019998357544230558,
      "loss": 2.7316,
      "step": 103
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.0625,
      "learning_rate": 0.00019997080140801932,
      "loss": 2.0909,
      "step": 104
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.0625,
      "learning_rate": 0.00019995437844895334,
      "loss": 2.6797,
      "step": 105
    },
    {
      "epoch": 0.17,
      "grad_norm": 23.375,
      "learning_rate": 0.00019993430716454413,
      "loss": 2.1525,
      "step": 106
    },
    {
      "epoch": 0.17,
      "grad_norm": 13.25,
      "learning_rate": 0.00019991058828739165,
      "loss": 2.6369,
      "step": 107
    },
    {
      "epoch": 0.17,
      "grad_norm": 11.0,
      "learning_rate": 0.00019988322268323268,
      "loss": 2.7311,
      "step": 108
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.9375,
      "learning_rate": 0.00019985221135090914,
      "loss": 2.3733,
      "step": 109
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.875,
      "learning_rate": 0.00019981755542233177,
      "loss": 2.2644,
      "step": 110
    },
    {
      "epoch": 0.18,
      "grad_norm": 12.4375,
      "learning_rate": 0.00019977925616243862,
      "loss": 2.5241,
      "step": 111
    },
    {
      "epoch": 0.18,
      "grad_norm": 15.875,
      "learning_rate": 0.00019973731496914914,
      "loss": 2.8353,
      "step": 112
    },
    {
      "epoch": 0.18,
      "grad_norm": 11.6875,
      "learning_rate": 0.0001996917333733128,
      "loss": 2.5239,
      "step": 113
    },
    {
      "epoch": 0.18,
      "grad_norm": 17.75,
      "learning_rate": 0.00019964251303865362,
      "loss": 3.1981,
      "step": 114
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.9375,
      "learning_rate": 0.00019958965576170908,
      "loss": 2.609,
      "step": 115
    },
    {
      "epoch": 0.19,
      "grad_norm": 98.0,
      "learning_rate": 0.00019953316347176488,
      "loss": 2.8462,
      "step": 116
    },
    {
      "epoch": 0.19,
      "grad_norm": 11.125,
      "learning_rate": 0.00019947303823078416,
      "loss": 2.6904,
      "step": 117
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.6875,
      "learning_rate": 0.00019940928223333252,
      "loss": 2.2545,
      "step": 118
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.1875,
      "learning_rate": 0.0001993418978064979,
      "loss": 2.6032,
      "step": 119
    },
    {
      "epoch": 0.19,
      "grad_norm": 13.25,
      "learning_rate": 0.0001992708874098054,
      "loss": 3.9633,
      "step": 120
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.25,
      "learning_rate": 0.00019919625363512786,
      "loss": 2.4602,
      "step": 121
    },
    {
      "epoch": 0.2,
      "grad_norm": 71.5,
      "learning_rate": 0.00019911799920659093,
      "loss": 2.213,
      "step": 122
    },
    {
      "epoch": 0.2,
      "grad_norm": 16.625,
      "learning_rate": 0.00019903612698047383,
      "loss": 2.9735,
      "step": 123
    },
    {
      "epoch": 0.2,
      "grad_norm": 19.5,
      "learning_rate": 0.0001989506399451051,
      "loss": 3.2784,
      "step": 124
    },
    {
      "epoch": 0.2,
      "grad_norm": 26.875,
      "learning_rate": 0.00019886154122075343,
      "loss": 2.0087,
      "step": 125
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.1875,
      "learning_rate": 0.00019876883405951377,
      "loss": 2.2212,
      "step": 126
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.375,
      "learning_rate": 0.00019867252184518878,
      "loss": 2.2147,
      "step": 127
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.625,
      "learning_rate": 0.0001985726080931651,
      "loss": 2.1791,
      "step": 128
    },
    {
      "epoch": 0.21,
      "grad_norm": 180.0,
      "learning_rate": 0.00019846909645028523,
      "loss": 2.6698,
      "step": 129
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.8125,
      "learning_rate": 0.00019836199069471437,
      "loss": 2.5204,
      "step": 130
    },
    {
      "epoch": 0.21,
      "grad_norm": 17.5,
      "learning_rate": 0.0001982512947358024,
      "loss": 2.1382,
      "step": 131
    },
    {
      "epoch": 0.21,
      "grad_norm": 13.6875,
      "learning_rate": 0.00019813701261394136,
      "loss": 2.9256,
      "step": 132
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.875,
      "learning_rate": 0.00019801914850041784,
      "loss": 2.2253,
      "step": 133
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.40625,
      "learning_rate": 0.00019789770669726087,
      "loss": 2.5591,
      "step": 134
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.0,
      "learning_rate": 0.00019777269163708468,
      "loss": 1.7038,
      "step": 135
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.5,
      "learning_rate": 0.00019764410788292722,
      "loss": 2.7535,
      "step": 136
    },
    {
      "epoch": 0.22,
      "grad_norm": 19.0,
      "learning_rate": 0.00019751196012808325,
      "loss": 2.727,
      "step": 137
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.9375,
      "learning_rate": 0.00019737625319593335,
      "loss": 2.671,
      "step": 138
    },
    {
      "epoch": 0.22,
      "grad_norm": 115.0,
      "learning_rate": 0.00019723699203976766,
      "loss": 9.1845,
      "step": 139
    },
    {
      "epoch": 0.23,
      "grad_norm": 59.0,
      "learning_rate": 0.0001970941817426052,
      "loss": 9.8542,
      "step": 140
    },
    {
      "epoch": 0.23,
      "grad_norm": 1632.0,
      "learning_rate": 0.00019694782751700828,
      "loss": 7.3106,
      "step": 141
    },
    {
      "epoch": 0.23,
      "grad_norm": 40.75,
      "learning_rate": 0.00019679793470489228,
      "loss": 9.6883,
      "step": 142
    },
    {
      "epoch": 0.23,
      "grad_norm": 816.0,
      "learning_rate": 0.00019664450877733062,
      "loss": 7.9838,
      "step": 143
    },
    {
      "epoch": 0.23,
      "grad_norm": 632.0,
      "learning_rate": 0.00019648755533435518,
      "loss": 4.5732,
      "step": 144
    },
    {
      "epoch": 0.23,
      "grad_norm": 580.0,
      "learning_rate": 0.00019632708010475165,
      "loss": 3.554,
      "step": 145
    },
    {
      "epoch": 0.24,
      "grad_norm": 872.0,
      "learning_rate": 0.00019616308894585078,
      "loss": 3.2136,
      "step": 146
    },
    {
      "epoch": 0.24,
      "grad_norm": 852.0,
      "learning_rate": 0.0001959955878433143,
      "loss": 4.1706,
      "step": 147
    },
    {
      "epoch": 0.24,
      "grad_norm": 204.0,
      "learning_rate": 0.00019582458291091663,
      "loss": 3.1317,
      "step": 148
    },
    {
      "epoch": 0.24,
      "grad_norm": 123.5,
      "learning_rate": 0.00019565008039032158,
      "loss": 3.9938,
      "step": 149
    },
    {
      "epoch": 0.24,
      "grad_norm": 1448.0,
      "learning_rate": 0.00019547208665085457,
      "loss": 7.1668,
      "step": 150
    },
    {
      "epoch": 0.24,
      "grad_norm": 2192.0,
      "learning_rate": 0.0001952906081892703,
      "loss": 5.1547,
      "step": 151
    },
    {
      "epoch": 0.25,
      "grad_norm": 39.0,
      "learning_rate": 0.00019510565162951537,
      "loss": 3.9528,
      "step": 152
    },
    {
      "epoch": 0.25,
      "grad_norm": 18.125,
      "learning_rate": 0.0001949172237224867,
      "loss": 3.2722,
      "step": 153
    },
    {
      "epoch": 0.25,
      "grad_norm": 1920.0,
      "learning_rate": 0.00019472533134578507,
      "loss": 8.9408,
      "step": 154
    },
    {
      "epoch": 0.25,
      "grad_norm": 2304.0,
      "learning_rate": 0.00019452998150346401,
      "loss": 9.7725,
      "step": 155
    },
    {
      "epoch": 0.25,
      "eval_loss": 8.848782539367676,
      "eval_runtime": 1.7872,
      "eval_samples_per_second": 19.024,
      "eval_steps_per_second": 19.024,
      "step": 155
    },
    {
      "epoch": 0.25,
      "grad_norm": 33.25,
      "learning_rate": 0.0001943311813257743,
      "loss": 9.3742,
      "step": 156
    },
    {
      "epoch": 0.25,
      "grad_norm": 232.0,
      "learning_rate": 0.00019412893806890357,
      "loss": 8.3919,
      "step": 157
    },
    {
      "epoch": 0.25,
      "grad_norm": 492.0,
      "learning_rate": 0.00019392325911471155,
      "loss": 7.3422,
      "step": 158
    },
    {
      "epoch": 0.26,
      "grad_norm": 872.0,
      "learning_rate": 0.00019371415197046052,
      "loss": 9.4552,
      "step": 159
    },
    {
      "epoch": 0.26,
      "grad_norm": 227.0,
      "learning_rate": 0.0001935016242685415,
      "loss": 3.9771,
      "step": 160
    },
    {
      "epoch": 0.26,
      "grad_norm": 116.0,
      "learning_rate": 0.00019328568376619543,
      "loss": 13.2183,
      "step": 161
    },
    {
      "epoch": 0.26,
      "grad_norm": 2720.0,
      "learning_rate": 0.00019306633834523024,
      "loss": 3.3559,
      "step": 162
    },
    {
      "epoch": 0.26,
      "grad_norm": 137.0,
      "learning_rate": 0.00019284359601173294,
      "loss": 3.1494,
      "step": 163
    },
    {
      "epoch": 0.26,
      "grad_norm": 161.0,
      "learning_rate": 0.00019261746489577765,
      "loss": 7.2183,
      "step": 164
    },
    {
      "epoch": 0.27,
      "grad_norm": 24.75,
      "learning_rate": 0.0001923879532511287,
      "loss": 8.7516,
      "step": 165
    },
    {
      "epoch": 0.27,
      "grad_norm": 856.0,
      "learning_rate": 0.0001921550694549393,
      "loss": 5.0249,
      "step": 166
    },
    {
      "epoch": 0.27,
      "grad_norm": 1096.0,
      "learning_rate": 0.000191918822007446,
      "loss": 5.1011,
      "step": 167
    },
    {
      "epoch": 0.27,
      "grad_norm": 760.0,
      "learning_rate": 0.00019167921953165825,
      "loss": 7.5649,
      "step": 168
    },
    {
      "epoch": 0.27,
      "grad_norm": 58.75,
      "learning_rate": 0.0001914362707730437,
      "loss": 3.2093,
      "step": 169
    },
    {
      "epoch": 0.27,
      "grad_norm": 208.0,
      "learning_rate": 0.00019118998459920902,
      "loss": 5.9444,
      "step": 170
    },
    {
      "epoch": 0.28,
      "grad_norm": 16.875,
      "learning_rate": 0.00019094036999957624,
      "loss": 8.3708,
      "step": 171
    },
    {
      "epoch": 0.28,
      "grad_norm": 324.0,
      "learning_rate": 0.00019068743608505455,
      "loss": 8.2079,
      "step": 172
    },
    {
      "epoch": 0.28,
      "grad_norm": 394.0,
      "learning_rate": 0.00019043119208770793,
      "loss": 7.2584,
      "step": 173
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.4375,
      "learning_rate": 0.00019017164736041795,
      "loss": 7.9704,
      "step": 174
    },
    {
      "epoch": 0.28,
      "grad_norm": 972.0,
      "learning_rate": 0.00018990881137654258,
      "loss": 9.457,
      "step": 175
    },
    {
      "epoch": 0.28,
      "grad_norm": 648.0,
      "learning_rate": 0.00018964269372957038,
      "loss": 5.1865,
      "step": 176
    },
    {
      "epoch": 0.29,
      "grad_norm": 12.25,
      "learning_rate": 0.0001893733041327702,
      "loss": 7.7786,
      "step": 177
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.875,
      "learning_rate": 0.0001891006524188368,
      "loss": 7.6235,
      "step": 178
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.5625,
      "learning_rate": 0.0001888247485395319,
      "loss": 7.4106,
      "step": 179
    },
    {
      "epoch": 0.29,
      "grad_norm": 13.0625,
      "learning_rate": 0.000188545602565321,
      "loss": 7.8207,
      "step": 180
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.75,
      "learning_rate": 0.00018826322468500566,
      "loss": 7.7207,
      "step": 181
    },
    {
      "epoch": 0.29,
      "grad_norm": 264.0,
      "learning_rate": 0.00018797762520535177,
      "loss": 6.7754,
      "step": 182
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.28125,
      "learning_rate": 0.00018768881455071332,
      "loss": 7.586,
      "step": 183
    },
    {
      "epoch": 0.3,
      "grad_norm": 143.0,
      "learning_rate": 0.0001873968032626518,
      "loss": 5.2382,
      "step": 184
    },
    {
      "epoch": 0.3,
      "grad_norm": 73.0,
      "learning_rate": 0.00018710160199955156,
      "loss": 3.6274,
      "step": 185
    },
    {
      "epoch": 0.3,
      "grad_norm": 11.625,
      "learning_rate": 0.00018680322153623075,
      "loss": 7.4444,
      "step": 186
    },
    {
      "epoch": 0.3,
      "grad_norm": 11.3125,
      "learning_rate": 0.000186501672763548,
      "loss": 7.4017,
      "step": 187
    },
    {
      "epoch": 0.3,
      "grad_norm": 125.0,
      "learning_rate": 0.00018619696668800492,
      "loss": 3.8111,
      "step": 188
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.5,
      "learning_rate": 0.00018588911443134448,
      "loss": 7.5166,
      "step": 189
    },
    {
      "epoch": 0.31,
      "grad_norm": 79.5,
      "learning_rate": 0.00018557812723014476,
      "loss": 7.4288,
      "step": 190
    },
    {
      "epoch": 0.31,
      "grad_norm": 398.0,
      "learning_rate": 0.00018526401643540922,
      "loss": 4.29,
      "step": 191
    },
    {
      "epoch": 0.31,
      "grad_norm": 180.0,
      "learning_rate": 0.0001849467935121521,
      "loss": 6.6381,
      "step": 192
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.875,
      "learning_rate": 0.00018462647003898006,
      "loss": 7.5981,
      "step": 193
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.5625,
      "learning_rate": 0.00018430305770766948,
      "loss": 7.3688,
      "step": 194
    },
    {
      "epoch": 0.31,
      "grad_norm": 500.0,
      "learning_rate": 0.0001839765683227398,
      "loss": 4.5837,
      "step": 195
    },
    {
      "epoch": 0.32,
      "grad_norm": 13504.0,
      "learning_rate": 0.00018364701380102266,
      "loss": 6.6357,
      "step": 196
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.09375,
      "learning_rate": 0.00018331440617122696,
      "loss": 7.4372,
      "step": 197
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.0,
      "learning_rate": 0.00018297875757349952,
      "loss": 7.3396,
      "step": 198
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.34375,
      "learning_rate": 0.00018264008025898248,
      "loss": 7.2063,
      "step": 199
    },
    {
      "epoch": 0.32,
      "grad_norm": 180.0,
      "learning_rate": 0.00018229838658936564,
      "loss": 6.5597,
      "step": 200
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.53125,
      "learning_rate": 0.00018195368903643563,
      "loss": 7.4723,
      "step": 201
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.8125,
      "learning_rate": 0.0001816060001816205,
      "loss": 7.2379,
      "step": 202
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.0,
      "learning_rate": 0.00018125533271553043,
      "loss": 7.4611,
      "step": 203
    },
    {
      "epoch": 0.33,
      "grad_norm": 86.0,
      "learning_rate": 0.00018090169943749476,
      "loss": 7.1709,
      "step": 204
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.59375,
      "learning_rate": 0.0001805451132550946,
      "loss": 7.2826,
      "step": 205
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.53125,
      "learning_rate": 0.00018018558718369186,
      "loss": 7.304,
      "step": 206
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.1875,
      "learning_rate": 0.00017982313434595406,
      "loss": 7.1153,
      "step": 207
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.625,
      "learning_rate": 0.00017945776797137543,
      "loss": 7.4778,
      "step": 208
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.1875,
      "learning_rate": 0.00017908950139579406,
      "loss": 7.2837,
      "step": 209
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.3125,
      "learning_rate": 0.00017871834806090501,
      "loss": 7.0416,
      "step": 210
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.78125,
      "learning_rate": 0.0001783443215137699,
      "loss": 7.2605,
      "step": 211
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.15625,
      "learning_rate": 0.00017796743540632223,
      "loss": 7.0665,
      "step": 212
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.90625,
      "learning_rate": 0.00017758770349486923,
      "loss": 7.1735,
      "step": 213
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.8125,
      "learning_rate": 0.00017720513963958968,
      "loss": 7.0647,
      "step": 214
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.5,
      "learning_rate": 0.00017681975780402807,
      "loss": 7.7996,
      "step": 215
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.15625,
      "learning_rate": 0.00017643157205458483,
      "loss": 7.4716,
      "step": 216
    },
    {
      "epoch": 0.35,
      "grad_norm": 14.6875,
      "learning_rate": 0.0001760405965600031,
      "loss": 7.1858,
      "step": 217
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.125,
      "learning_rate": 0.00017564684559085136,
      "loss": 7.2222,
      "step": 218
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.3125,
      "learning_rate": 0.00017525033351900268,
      "loss": 7.4119,
      "step": 219
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.25,
      "learning_rate": 0.00017485107481711012,
      "loss": 7.3965,
      "step": 220
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.125,
      "learning_rate": 0.00017444908405807845,
      "loss": 7.2064,
      "step": 221
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.5,
      "learning_rate": 0.00017404437591453235,
      "loss": 7.6118,
      "step": 222
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.6875,
      "learning_rate": 0.00017363696515828062,
      "loss": 7.1017,
      "step": 223
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.0625,
      "learning_rate": 0.00017322686665977737,
      "loss": 6.9061,
      "step": 224
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.5,
      "learning_rate": 0.00017281409538757883,
      "loss": 7.2821,
      "step": 225
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.0,
      "learning_rate": 0.00017239866640779745,
      "loss": 7.6545,
      "step": 226
    },
    {
      "epoch": 0.37,
      "grad_norm": 12.125,
      "learning_rate": 0.0001719805948835515,
      "loss": 7.4611,
      "step": 227
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.875,
      "learning_rate": 0.00017155989607441213,
      "loss": 6.9964,
      "step": 228
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.4375,
      "learning_rate": 0.00017113658533584594,
      "loss": 7.3558,
      "step": 229
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.8125,
      "learning_rate": 0.00017071067811865476,
      "loss": 7.2276,
      "step": 230
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.625,
      "learning_rate": 0.00017028218996841172,
      "loss": 7.2778,
      "step": 231
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.28125,
      "learning_rate": 0.00016985113652489374,
      "loss": 7.1171,
      "step": 232
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.90625,
      "learning_rate": 0.00016941753352151055,
      "loss": 7.5252,
      "step": 233
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.71875,
      "learning_rate": 0.00016898139678473076,
      "loss": 7.3942,
      "step": 234
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.375,
      "learning_rate": 0.00016854274223350397,
      "loss": 7.2534,
      "step": 235
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.65625,
      "learning_rate": 0.00016810158587867973,
      "loss": 7.3688,
      "step": 236
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.96875,
      "learning_rate": 0.00016765794382242314,
      "loss": 7.3569,
      "step": 237
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.34375,
      "learning_rate": 0.00016721183225762727,
      "loss": 6.9082,
      "step": 238
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.5,
      "learning_rate": 0.00016676326746732195,
      "loss": 7.2176,
      "step": 239
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.75,
      "learning_rate": 0.00016631226582407952,
      "loss": 6.8296,
      "step": 240
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.4375,
      "learning_rate": 0.00016585884378941725,
      "loss": 7.4089,
      "step": 241
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.9375,
      "learning_rate": 0.00016540301791319645,
      "loss": 7.2357,
      "step": 242
    },
    {
      "epoch": 0.39,
      "grad_norm": 4.84375,
      "learning_rate": 0.00016494480483301836,
      "loss": 7.3099,
      "step": 243
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.1875,
      "learning_rate": 0.00016448422127361706,
      "loss": 6.7821,
      "step": 244
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.03125,
      "learning_rate": 0.00016402128404624882,
      "loss": 7.4242,
      "step": 245
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.84375,
      "learning_rate": 0.00016355601004807856,
      "loss": 7.0436,
      "step": 246
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.46875,
      "learning_rate": 0.00016308841626156307,
      "loss": 7.2183,
      "step": 247
    },
    {
      "epoch": 0.4,
      "grad_norm": 13.6875,
      "learning_rate": 0.00016261851975383137,
      "loss": 7.2762,
      "step": 248
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.25,
      "learning_rate": 0.00016214633767606143,
      "loss": 7.6098,
      "step": 249
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.3125,
      "learning_rate": 0.00016167188726285434,
      "loss": 7.2284,
      "step": 250
    },
    {
      "epoch": 0.4,
      "grad_norm": 18.625,
      "learning_rate": 0.0001611951858316052,
      "loss": 7.4437,
      "step": 251
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.34375,
      "learning_rate": 0.00016071625078187114,
      "loss": 7.2706,
      "step": 252
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.375,
      "learning_rate": 0.00016023509959473605,
      "loss": 7.1823,
      "step": 253
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.8125,
      "learning_rate": 0.00015975174983217275,
      "loss": 7.4077,
      "step": 254
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.90625,
      "learning_rate": 0.0001592662191364017,
      "loss": 7.3115,
      "step": 255
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.375,
      "learning_rate": 0.00015877852522924732,
      "loss": 7.6855,
      "step": 256
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.5625,
      "learning_rate": 0.00015828868591149104,
      "loss": 7.3519,
      "step": 257
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.90625,
      "learning_rate": 0.0001577967190622215,
      "loss": 7.4478,
      "step": 258
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.59375,
      "learning_rate": 0.00015730264263818212,
      "loss": 7.4721,
      "step": 259
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.375,
      "learning_rate": 0.00015680647467311557,
      "loss": 7.25,
      "step": 260
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.28125,
      "learning_rate": 0.00015630823327710558,
      "loss": 7.278,
      "step": 261
    },
    {
      "epoch": 0.42,
      "grad_norm": 932.0,
      "learning_rate": 0.00015580793663591585,
      "loss": 7.8209,
      "step": 262
    },
    {
      "epoch": 0.42,
      "grad_norm": 13.4375,
      "learning_rate": 0.0001553056030103264,
      "loss": 7.4705,
      "step": 263
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.34375,
      "learning_rate": 0.00015480125073546704,
      "loss": 7.2755,
      "step": 264
    },
    {
      "epoch": 0.43,
      "grad_norm": 628.0,
      "learning_rate": 0.0001542948982201479,
      "loss": 7.3964,
      "step": 265
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.4375,
      "learning_rate": 0.00015378656394618787,
      "loss": 7.259,
      "step": 266
    },
    {
      "epoch": 0.43,
      "grad_norm": 29.25,
      "learning_rate": 0.00015327626646773976,
      "loss": 8.2839,
      "step": 267
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.78125,
      "learning_rate": 0.0001527640244106133,
      "loss": 7.4219,
      "step": 268
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.21875,
      "learning_rate": 0.0001522498564715949,
      "loss": 7.4376,
      "step": 269
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.875,
      "learning_rate": 0.00015173378141776568,
      "loss": 7.3749,
      "step": 270
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.21875,
      "learning_rate": 0.00015121581808581622,
      "loss": 7.1775,
      "step": 271
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.4375,
      "learning_rate": 0.00015069598538135906,
      "loss": 7.2296,
      "step": 272
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.46875,
      "learning_rate": 0.00015017430227823864,
      "loss": 7.3572,
      "step": 273
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.8125,
      "learning_rate": 0.0001496507878178388,
      "loss": 7.2276,
      "step": 274
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.75,
      "learning_rate": 0.00014912546110838775,
      "loss": 7.5138,
      "step": 275
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.125,
      "learning_rate": 0.0001485983413242606,
      "loss": 7.0377,
      "step": 276
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.65625,
      "learning_rate": 0.00014806944770527958,
      "loss": 7.1247,
      "step": 277
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.1875,
      "learning_rate": 0.00014753879955601163,
      "loss": 7.3482,
      "step": 278
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.5,
      "learning_rate": 0.00014700641624506392,
      "loss": 6.9676,
      "step": 279
    },
    {
      "epoch": 0.45,
      "grad_norm": 6.03125,
      "learning_rate": 0.00014647231720437686,
      "loss": 7.2884,
      "step": 280
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.0,
      "learning_rate": 0.00014593652192851486,
      "loss": 7.558,
      "step": 281
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.65625,
      "learning_rate": 0.00014539904997395468,
      "loss": 7.1275,
      "step": 282
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.875,
      "learning_rate": 0.00014485992095837177,
      "loss": 6.9928,
      "step": 283
    },
    {
      "epoch": 0.46,
      "grad_norm": 14.6875,
      "learning_rate": 0.00014431915455992414,
      "loss": 7.2045,
      "step": 284
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.96875,
      "learning_rate": 0.00014377677051653404,
      "loss": 7.1944,
      "step": 285
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.5,
      "learning_rate": 0.00014323278862516775,
      "loss": 7.8461,
      "step": 286
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.40625,
      "learning_rate": 0.00014268722874111265,
      "loss": 7.1609,
      "step": 287
    },
    {
      "epoch": 0.46,
      "grad_norm": 1232.0,
      "learning_rate": 0.00014214011077725292,
      "loss": 8.6244,
      "step": 288
    },
    {
      "epoch": 0.47,
      "grad_norm": 422.0,
      "learning_rate": 0.00014159145470334235,
      "loss": 8.2071,
      "step": 289
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.28125,
      "learning_rate": 0.0001410412805452757,
      "loss": 7.1693,
      "step": 290
    },
    {
      "epoch": 0.47,
      "grad_norm": 24.25,
      "learning_rate": 0.00014048960838435753,
      "loss": 7.1325,
      "step": 291
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.5,
      "learning_rate": 0.00013993645835656953,
      "loss": 7.2788,
      "step": 292
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.5625,
      "learning_rate": 0.00013938185065183532,
      "loss": 7.3263,
      "step": 293
    },
    {
      "epoch": 0.47,
      "grad_norm": 72.5,
      "learning_rate": 0.0001388258055132835,
      "loss": 7.4291,
      "step": 294
    },
    {
      "epoch": 0.48,
      "grad_norm": 11.125,
      "learning_rate": 0.000138268343236509,
      "loss": 7.3335,
      "step": 295
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.09375,
      "learning_rate": 0.00013770948416883205,
      "loss": 7.4262,
      "step": 296
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.5,
      "learning_rate": 0.00013714924870855571,
      "loss": 7.3771,
      "step": 297
    },
    {
      "epoch": 0.48,
      "grad_norm": 16.375,
      "learning_rate": 0.00013658765730422125,
      "loss": 7.5395,
      "step": 298
    },
    {
      "epoch": 0.48,
      "grad_norm": 15.0625,
      "learning_rate": 0.00013602473045386165,
      "loss": 7.6106,
      "step": 299
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.90625,
      "learning_rate": 0.00013546048870425356,
      "loss": 7.3483,
      "step": 300
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.5625,
      "learning_rate": 0.0001348949526501675,
      "loss": 8.0617,
      "step": 301
    },
    {
      "epoch": 0.49,
      "grad_norm": 25.25,
      "learning_rate": 0.00013432814293361584,
      "loss": 6.8243,
      "step": 302
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.8125,
      "learning_rate": 0.00013376008024309948,
      "loss": 7.3006,
      "step": 303
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.375,
      "learning_rate": 0.00013319078531285285,
      "loss": 7.3865,
      "step": 304
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.0,
      "learning_rate": 0.00013262027892208694,
      "loss": 7.676,
      "step": 305
    },
    {
      "epoch": 0.49,
      "grad_norm": 13.75,
      "learning_rate": 0.00013204858189423097,
      "loss": 7.3239,
      "step": 306
    },
    {
      "epoch": 0.5,
      "grad_norm": 3120.0,
      "learning_rate": 0.00013147571509617228,
      "loss": 8.0707,
      "step": 307
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.96875,
      "learning_rate": 0.00013090169943749476,
      "loss": 7.0195,
      "step": 308
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.1875,
      "learning_rate": 0.00013032655586971552,
      "loss": 7.367,
      "step": 309
    },
    {
      "epoch": 0.5,
      "grad_norm": 18.5,
      "learning_rate": 0.00012975030538552032,
      "loss": 7.687,
      "step": 310
    },
    {
      "epoch": 0.5,
      "eval_loss": 7.375539302825928,
      "eval_runtime": 1.796,
      "eval_samples_per_second": 18.931,
      "eval_steps_per_second": 18.931,
      "step": 310
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.0,
      "learning_rate": 0.0001291729690179972,
      "loss": 7.4606,
      "step": 311
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.375,
      "learning_rate": 0.00012859456783986893,
      "loss": 7.5164,
      "step": 312
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.65625,
      "learning_rate": 0.00012801512296272368,
      "loss": 7.685,
      "step": 313
    },
    {
      "epoch": 0.51,
      "grad_norm": 6.5625,
      "learning_rate": 0.0001274346555362446,
      "loss": 7.2255,
      "step": 314
    },
    {
      "epoch": 0.51,
      "grad_norm": 17.125,
      "learning_rate": 0.0001268531867474377,
      "loss": 7.8333,
      "step": 315
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.125,
      "learning_rate": 0.0001262707378198587,
      "loss": 7.4813,
      "step": 316
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.375,
      "learning_rate": 0.00012568733001283827,
      "loss": 7.2433,
      "step": 317
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.90625,
      "learning_rate": 0.00012510298462070619,
      "loss": 7.4513,
      "step": 318
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.53125,
      "learning_rate": 0.00012451772297201376,
      "loss": 7.31,
      "step": 319
    },
    {
      "epoch": 0.52,
      "grad_norm": 30.0,
      "learning_rate": 0.0001239315664287558,
      "loss": 7.9245,
      "step": 320
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.03125,
      "learning_rate": 0.00012334453638559057,
      "loss": 7.2468,
      "step": 321
    },
    {
      "epoch": 0.52,
      "grad_norm": 28.5,
      "learning_rate": 0.000122756654269059,
      "loss": 8.0352,
      "step": 322
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.34375,
      "learning_rate": 0.00012216794153680274,
      "loss": 7.2967,
      "step": 323
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.375,
      "learning_rate": 0.00012157841967678063,
      "loss": 7.2094,
      "step": 324
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.4375,
      "learning_rate": 0.00012098811020648475,
      "loss": 7.2978,
      "step": 325
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.625,
      "learning_rate": 0.00012039703467215488,
      "loss": 7.3662,
      "step": 326
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.28125,
      "learning_rate": 0.00011980521464799198,
      "loss": 7.2527,
      "step": 327
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.5625,
      "learning_rate": 0.00011921267173537086,
      "loss": 7.616,
      "step": 328
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.0625,
      "learning_rate": 0.00011861942756205169,
      "loss": 7.1239,
      "step": 329
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.78125,
      "learning_rate": 0.0001180255037813906,
      "loss": 7.1963,
      "step": 330
    },
    {
      "epoch": 0.53,
      "grad_norm": 47.0,
      "learning_rate": 0.00011743092207154929,
      "loss": 7.1912,
      "step": 331
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.25,
      "learning_rate": 0.00011683570413470383,
      "loss": 7.3671,
      "step": 332
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.09375,
      "learning_rate": 0.00011623987169625261,
      "loss": 7.2449,
      "step": 333
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.0625,
      "learning_rate": 0.0001156434465040231,
      "loss": 7.3294,
      "step": 334
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.375,
      "learning_rate": 0.00011504645032747832,
      "loss": 7.3717,
      "step": 335
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.65625,
      "learning_rate": 0.00011444890495692213,
      "loss": 7.1599,
      "step": 336
    },
    {
      "epoch": 0.54,
      "grad_norm": 28.25,
      "learning_rate": 0.00011385083220270401,
      "loss": 7.4863,
      "step": 337
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.78125,
      "learning_rate": 0.00011325225389442277,
      "loss": 7.7702,
      "step": 338
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.3125,
      "learning_rate": 0.00011265319188012994,
      "loss": 7.0122,
      "step": 339
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.15625,
      "learning_rate": 0.0001120536680255323,
      "loss": 7.4499,
      "step": 340
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.4375,
      "learning_rate": 0.00011145370421319377,
      "loss": 7.1818,
      "step": 341
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.09375,
      "learning_rate": 0.00011085332234173664,
      "loss": 7.589,
      "step": 342
    },
    {
      "epoch": 0.55,
      "grad_norm": 12.3125,
      "learning_rate": 0.00011025254432504233,
      "loss": 7.8762,
      "step": 343
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.5625,
      "learning_rate": 0.00010965139209145152,
      "loss": 7.3293,
      "step": 344
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.71875,
      "learning_rate": 0.0001090498875829638,
      "loss": 7.4994,
      "step": 345
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.625,
      "learning_rate": 0.00010844805275443673,
      "loss": 7.185,
      "step": 346
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.6875,
      "learning_rate": 0.0001078459095727845,
      "loss": 7.1749,
      "step": 347
    },
    {
      "epoch": 0.56,
      "grad_norm": 17.25,
      "learning_rate": 0.00010724348001617625,
      "loss": 7.6093,
      "step": 348
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.0625,
      "learning_rate": 0.00010664078607323367,
      "loss": 7.1092,
      "step": 349
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.34375,
      "learning_rate": 0.00010603784974222861,
      "loss": 7.2545,
      "step": 350
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.28125,
      "learning_rate": 0.00010543469303028002,
      "loss": 7.2538,
      "step": 351
    },
    {
      "epoch": 0.57,
      "grad_norm": 17.25,
      "learning_rate": 0.00010483133795255071,
      "loss": 7.9843,
      "step": 352
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.875,
      "learning_rate": 0.0001042278065314439,
      "loss": 7.931,
      "step": 353
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.5,
      "learning_rate": 0.00010362412079579924,
      "loss": 7.1528,
      "step": 354
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.875,
      "learning_rate": 0.0001030203027800889,
      "loss": 7.2796,
      "step": 355
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.84375,
      "learning_rate": 0.00010241637452361323,
      "loss": 7.4538,
      "step": 356
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.15625,
      "learning_rate": 0.0001018123580696964,
      "loss": 7.3754,
      "step": 357
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.03125,
      "learning_rate": 0.00010120827546488174,
      "loss": 7.5391,
      "step": 358
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.375,
      "learning_rate": 0.00010060414875812709,
      "loss": 7.0525,
      "step": 359
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.28125,
      "learning_rate": 0.0001,
      "loss": 7.2558,
      "step": 360
    },
    {
      "epoch": 0.58,
      "grad_norm": 11.6875,
      "learning_rate": 9.939585124187292e-05,
      "loss": 7.33,
      "step": 361
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.5625,
      "learning_rate": 9.879172453511827e-05,
      "loss": 7.0643,
      "step": 362
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.90625,
      "learning_rate": 9.818764193030363e-05,
      "loss": 7.3601,
      "step": 363
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.1875,
      "learning_rate": 9.75836254763868e-05,
      "loss": 7.1155,
      "step": 364
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.125,
      "learning_rate": 9.697969721991114e-05,
      "loss": 7.1084,
      "step": 365
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.8125,
      "learning_rate": 9.63758792042008e-05,
      "loss": 7.2947,
      "step": 366
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.1875,
      "learning_rate": 9.577219346855613e-05,
      "loss": 7.1426,
      "step": 367
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.8125,
      "learning_rate": 9.516866204744931e-05,
      "loss": 7.317,
      "step": 368
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.0625,
      "learning_rate": 9.456530696971999e-05,
      "loss": 7.3881,
      "step": 369
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.625,
      "learning_rate": 9.396215025777139e-05,
      "loss": 7.28,
      "step": 370
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.40625,
      "learning_rate": 9.335921392676631e-05,
      "loss": 7.2141,
      "step": 371
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.25,
      "learning_rate": 9.275651998382377e-05,
      "loss": 7.2653,
      "step": 372
    },
    {
      "epoch": 0.6,
      "grad_norm": 81.5,
      "learning_rate": 9.215409042721552e-05,
      "loss": 7.1614,
      "step": 373
    },
    {
      "epoch": 0.6,
      "grad_norm": 3.65625,
      "learning_rate": 9.155194724556331e-05,
      "loss": 7.1929,
      "step": 374
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.1875,
      "learning_rate": 9.095011241703623e-05,
      "loss": 7.0878,
      "step": 375
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.5,
      "learning_rate": 9.034860790854849e-05,
      "loss": 7.4207,
      "step": 376
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.65625,
      "learning_rate": 8.974745567495768e-05,
      "loss": 7.1018,
      "step": 377
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.65625,
      "learning_rate": 8.914667765826338e-05,
      "loss": 7.3069,
      "step": 378
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.125,
      "learning_rate": 8.854629578680624e-05,
      "loss": 7.4862,
      "step": 379
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.71875,
      "learning_rate": 8.79463319744677e-05,
      "loss": 7.2446,
      "step": 380
    },
    {
      "epoch": 0.61,
      "grad_norm": 188.0,
      "learning_rate": 8.73468081198701e-05,
      "loss": 7.2829,
      "step": 381
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.5625,
      "learning_rate": 8.674774610557728e-05,
      "loss": 7.3516,
      "step": 382
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.78125,
      "learning_rate": 8.614916779729603e-05,
      "loss": 7.1598,
      "step": 383
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.65625,
      "learning_rate": 8.55510950430779e-05,
      "loss": 7.307,
      "step": 384
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.5,
      "learning_rate": 8.495354967252169e-05,
      "loss": 7.3374,
      "step": 385
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.75,
      "learning_rate": 8.435655349597689e-05,
      "loss": 7.5147,
      "step": 386
    },
    {
      "epoch": 0.62,
      "grad_norm": 3.9375,
      "learning_rate": 8.37601283037474e-05,
      "loss": 7.6806,
      "step": 387
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.1875,
      "learning_rate": 8.316429586529615e-05,
      "loss": 7.2063,
      "step": 388
    },
    {
      "epoch": 0.63,
      "grad_norm": 3.703125,
      "learning_rate": 8.256907792845072e-05,
      "loss": 7.2237,
      "step": 389
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.375,
      "learning_rate": 8.197449621860943e-05,
      "loss": 6.9902,
      "step": 390
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.5,
      "learning_rate": 8.138057243794833e-05,
      "loss": 7.2775,
      "step": 391
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.65625,
      "learning_rate": 8.078732826462915e-05,
      "loss": 7.3332,
      "step": 392
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.71875,
      "learning_rate": 8.019478535200806e-05,
      "loss": 7.3288,
      "step": 393
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.703125,
      "learning_rate": 7.960296532784515e-05,
      "loss": 7.0595,
      "step": 394
    },
    {
      "epoch": 0.64,
      "grad_norm": 108.0,
      "learning_rate": 7.901188979351526e-05,
      "loss": 7.1364,
      "step": 395
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.9375,
      "learning_rate": 7.84215803232194e-05,
      "loss": 7.3006,
      "step": 396
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.34375,
      "learning_rate": 7.78320584631973e-05,
      "loss": 7.1831,
      "step": 397
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.6875,
      "learning_rate": 7.7243345730941e-05,
      "loss": 7.1149,
      "step": 398
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.65625,
      "learning_rate": 7.66554636144095e-05,
      "loss": 7.1149,
      "step": 399
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.375,
      "learning_rate": 7.606843357124426e-05,
      "loss": 7.2536,
      "step": 400
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.5625,
      "learning_rate": 7.548227702798624e-05,
      "loss": 7.1493,
      "step": 401
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.5,
      "learning_rate": 7.489701537929384e-05,
      "loss": 7.1997,
      "step": 402
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.25,
      "learning_rate": 7.431266998716171e-05,
      "loss": 7.3595,
      "step": 403
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.0,
      "learning_rate": 7.372926218014131e-05,
      "loss": 7.2021,
      "step": 404
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.1875,
      "learning_rate": 7.314681325256232e-05,
      "loss": 7.068,
      "step": 405
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.0625,
      "learning_rate": 7.256534446375542e-05,
      "loss": 7.0883,
      "step": 406
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.8125,
      "learning_rate": 7.198487703727632e-05,
      "loss": 7.1549,
      "step": 407
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.8125,
      "learning_rate": 7.14054321601311e-05,
      "loss": 7.2792,
      "step": 408
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.625,
      "learning_rate": 7.082703098200282e-05,
      "loss": 6.8635,
      "step": 409
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.1875,
      "learning_rate": 7.024969461447972e-05,
      "loss": 7.2292,
      "step": 410
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.3125,
      "learning_rate": 6.967344413028452e-05,
      "loss": 7.1733,
      "step": 411
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.9375,
      "learning_rate": 6.909830056250527e-05,
      "loss": 7.4127,
      "step": 412
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.0,
      "learning_rate": 6.852428490382773e-05,
      "loss": 7.2303,
      "step": 413
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.1875,
      "learning_rate": 6.795141810576906e-05,
      "loss": 7.0902,
      "step": 414
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.5625,
      "learning_rate": 6.73797210779131e-05,
      "loss": 7.3092,
      "step": 415
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.3125,
      "learning_rate": 6.680921468714719e-05,
      "loss": 7.0385,
      "step": 416
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.15625,
      "learning_rate": 6.623991975690051e-05,
      "loss": 7.2035,
      "step": 417
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.8125,
      "learning_rate": 6.567185706638417e-05,
      "loss": 7.327,
      "step": 418
    },
    {
      "epoch": 0.68,
      "grad_norm": 6.875,
      "learning_rate": 6.510504734983249e-05,
      "loss": 7.2225,
      "step": 419
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.3125,
      "learning_rate": 6.453951129574644e-05,
      "loss": 6.8114,
      "step": 420
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.84375,
      "learning_rate": 6.397526954613839e-05,
      "loss": 7.0789,
      "step": 421
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.40625,
      "learning_rate": 6.341234269577879e-05,
      "loss": 7.179,
      "step": 422
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.6875,
      "learning_rate": 6.285075129144428e-05,
      "loss": 7.2077,
      "step": 423
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.875,
      "learning_rate": 6.229051583116796e-05,
      "loss": 7.2929,
      "step": 424
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.3125,
      "learning_rate": 6.173165676349103e-05,
      "loss": 7.2718,
      "step": 425
    },
    {
      "epoch": 0.69,
      "grad_norm": 3.625,
      "learning_rate": 6.117419448671651e-05,
      "loss": 7.2209,
      "step": 426
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.0625,
      "learning_rate": 6.0618149348164696e-05,
      "loss": 7.1679,
      "step": 427
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.0625,
      "learning_rate": 6.006354164343046e-05,
      "loss": 7.1983,
      "step": 428
    },
    {
      "epoch": 0.69,
      "grad_norm": 3.6875,
      "learning_rate": 5.9510391615642466e-05,
      "loss": 7.1982,
      "step": 429
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.28125,
      "learning_rate": 5.8958719454724346e-05,
      "loss": 7.2756,
      "step": 430
    },
    {
      "epoch": 0.7,
      "grad_norm": 6.53125,
      "learning_rate": 5.840854529665767e-05,
      "loss": 7.0365,
      "step": 431
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.625,
      "learning_rate": 5.785988922274711e-05,
      "loss": 7.34,
      "step": 432
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.53125,
      "learning_rate": 5.7312771258887386e-05,
      "loss": 7.1793,
      "step": 433
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.5,
      "learning_rate": 5.676721137483225e-05,
      "loss": 7.227,
      "step": 434
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.25,
      "learning_rate": 5.622322948346594e-05,
      "loss": 7.2284,
      "step": 435
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.21875,
      "learning_rate": 5.568084544007588e-05,
      "loss": 7.6541,
      "step": 436
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.625,
      "learning_rate": 5.5140079041628214e-05,
      "loss": 7.5761,
      "step": 437
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.3125,
      "learning_rate": 5.4600950026045326e-05,
      "loss": 7.1913,
      "step": 438
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.03125,
      "learning_rate": 5.406347807148515e-05,
      "loss": 7.2958,
      "step": 439
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.65625,
      "learning_rate": 5.3527682795623146e-05,
      "loss": 7.2452,
      "step": 440
    },
    {
      "epoch": 0.71,
      "grad_norm": 6.15625,
      "learning_rate": 5.2993583754936126e-05,
      "loss": 7.1716,
      "step": 441
    },
    {
      "epoch": 0.71,
      "grad_norm": 27.75,
      "learning_rate": 5.246120044398839e-05,
      "loss": 7.2082,
      "step": 442
    },
    {
      "epoch": 0.71,
      "grad_norm": 6.8125,
      "learning_rate": 5.193055229472045e-05,
      "loss": 7.09,
      "step": 443
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.5,
      "learning_rate": 5.14016586757394e-05,
      "loss": 7.2108,
      "step": 444
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.25,
      "learning_rate": 5.087453889161229e-05,
      "loss": 7.3688,
      "step": 445
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.75,
      "learning_rate": 5.0349212182161254e-05,
      "loss": 7.3458,
      "step": 446
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.875,
      "learning_rate": 4.98256977217614e-05,
      "loss": 7.0711,
      "step": 447
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.5,
      "learning_rate": 4.9304014618640995e-05,
      "loss": 7.1625,
      "step": 448
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.15625,
      "learning_rate": 4.87841819141838e-05,
      "loss": 6.8236,
      "step": 449
    },
    {
      "epoch": 0.73,
      "grad_norm": 53.75,
      "learning_rate": 4.826621858223431e-05,
      "loss": 7.1651,
      "step": 450
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.0,
      "learning_rate": 4.7750143528405126e-05,
      "loss": 7.2596,
      "step": 451
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.21875,
      "learning_rate": 4.723597558938672e-05,
      "loss": 7.1167,
      "step": 452
    },
    {
      "epoch": 0.73,
      "grad_norm": 22.25,
      "learning_rate": 4.672373353226023e-05,
      "loss": 7.1937,
      "step": 453
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.25,
      "learning_rate": 4.6213436053812144e-05,
      "loss": 7.1933,
      "step": 454
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.59375,
      "learning_rate": 4.5705101779852135e-05,
      "loss": 7.4558,
      "step": 455
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.125,
      "learning_rate": 4.519874926453302e-05,
      "loss": 6.9894,
      "step": 456
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.40625,
      "learning_rate": 4.469439698967359e-05,
      "loss": 7.0223,
      "step": 457
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.5625,
      "learning_rate": 4.419206336408418e-05,
      "loss": 7.1284,
      "step": 458
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.0625,
      "learning_rate": 4.3691766722894435e-05,
      "loss": 7.3841,
      "step": 459
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.21875,
      "learning_rate": 4.3193525326884435e-05,
      "loss": 7.3895,
      "step": 460
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.5625,
      "learning_rate": 4.26973573618179e-05,
      "loss": 7.3107,
      "step": 461
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.75,
      "learning_rate": 4.220328093777851e-05,
      "loss": 7.2527,
      "step": 462
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.6875,
      "learning_rate": 4.1711314088509e-05,
      "loss": 7.1957,
      "step": 463
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.0625,
      "learning_rate": 4.12214747707527e-05,
      "loss": 7.1818,
      "step": 464
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.59375,
      "learning_rate": 4.0733780863598335e-05,
      "loss": 7.286,
      "step": 465
    },
    {
      "epoch": 0.75,
      "eval_loss": 7.306603908538818,
      "eval_runtime": 1.801,
      "eval_samples_per_second": 18.878,
      "eval_steps_per_second": 18.878,
      "step": 465
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.6875,
      "learning_rate": 4.0248250167827275e-05,
      "loss": 7.279,
      "step": 466
    },
    {
      "epoch": 0.75,
      "grad_norm": 26.125,
      "learning_rate": 3.976490040526394e-05,
      "loss": 7.1727,
      "step": 467
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.5625,
      "learning_rate": 3.9283749218128885e-05,
      "loss": 7.3203,
      "step": 468
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.46875,
      "learning_rate": 3.88048141683948e-05,
      "loss": 7.1131,
      "step": 469
    },
    {
      "epoch": 0.76,
      "grad_norm": 6.96875,
      "learning_rate": 3.832811273714569e-05,
      "loss": 7.4104,
      "step": 470
    },
    {
      "epoch": 0.76,
      "grad_norm": 6.09375,
      "learning_rate": 3.785366232393861e-05,
      "loss": 7.0627,
      "step": 471
    },
    {
      "epoch": 0.76,
      "grad_norm": 88.5,
      "learning_rate": 3.738148024616863e-05,
      "loss": 7.2023,
      "step": 472
    },
    {
      "epoch": 0.76,
      "grad_norm": 76.0,
      "learning_rate": 3.691158373843694e-05,
      "loss": 7.2856,
      "step": 473
    },
    {
      "epoch": 0.76,
      "grad_norm": 6.46875,
      "learning_rate": 3.644398995192147e-05,
      "loss": 7.5069,
      "step": 474
    },
    {
      "epoch": 0.77,
      "grad_norm": 4.28125,
      "learning_rate": 3.597871595375121e-05,
      "loss": 7.2348,
      "step": 475
    },
    {
      "epoch": 0.77,
      "grad_norm": 6.15625,
      "learning_rate": 3.5515778726382966e-05,
      "loss": 7.3896,
      "step": 476
    },
    {
      "epoch": 0.77,
      "grad_norm": 7.9375,
      "learning_rate": 3.5055195166981645e-05,
      "loss": 7.3101,
      "step": 477
    },
    {
      "epoch": 0.77,
      "grad_norm": 6.1875,
      "learning_rate": 3.459698208680359e-05,
      "loss": 7.4878,
      "step": 478
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.25,
      "learning_rate": 3.4141156210582756e-05,
      "loss": 7.7337,
      "step": 479
    },
    {
      "epoch": 0.77,
      "grad_norm": 7.71875,
      "learning_rate": 3.36877341759205e-05,
      "loss": 7.576,
      "step": 480
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.28125,
      "learning_rate": 3.3236732532678096e-05,
      "loss": 7.3218,
      "step": 481
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.125,
      "learning_rate": 3.2788167742372725e-05,
      "loss": 7.2678,
      "step": 482
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.8125,
      "learning_rate": 3.234205617757686e-05,
      "loss": 7.0266,
      "step": 483
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.75,
      "learning_rate": 3.1898414121320276e-05,
      "loss": 6.8119,
      "step": 484
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.9375,
      "learning_rate": 3.1457257766496015e-05,
      "loss": 7.1597,
      "step": 485
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.71875,
      "learning_rate": 3.101860321526924e-05,
      "loss": 7.6388,
      "step": 486
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.125,
      "learning_rate": 3.0582466478489455e-05,
      "loss": 7.367,
      "step": 487
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.0625,
      "learning_rate": 3.0148863475106314e-05,
      "loss": 7.2989,
      "step": 488
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.8125,
      "learning_rate": 2.9717810031588277e-05,
      "loss": 7.2247,
      "step": 489
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.6875,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 7.1624,
      "step": 490
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.375,
      "learning_rate": 2.886341466415412e-05,
      "loss": 7.1773,
      "step": 491
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.375,
      "learning_rate": 2.84401039255879e-05,
      "loss": 7.2401,
      "step": 492
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.125,
      "learning_rate": 2.8019405116448516e-05,
      "loss": 7.462,
      "step": 493
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.1875,
      "learning_rate": 2.7601333592202583e-05,
      "loss": 6.8809,
      "step": 494
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.1875,
      "learning_rate": 2.7185904612421176e-05,
      "loss": 7.1306,
      "step": 495
    },
    {
      "epoch": 0.8,
      "grad_norm": 6144.0,
      "learning_rate": 2.677313334022268e-05,
      "loss": 7.896,
      "step": 496
    },
    {
      "epoch": 0.8,
      "grad_norm": 6.40625,
      "learning_rate": 2.6363034841719392e-05,
      "loss": 7.2291,
      "step": 497
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.6875,
      "learning_rate": 2.59556240854677e-05,
      "loss": 7.538,
      "step": 498
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.1875,
      "learning_rate": 2.5550915941921526e-05,
      "loss": 7.2563,
      "step": 499
    },
    {
      "epoch": 0.81,
      "grad_norm": 588.0,
      "learning_rate": 2.514892518288988e-05,
      "loss": 7.2489,
      "step": 500
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.1875,
      "learning_rate": 2.4749666480997337e-05,
      "loss": 6.9217,
      "step": 501
    },
    {
      "epoch": 0.81,
      "grad_norm": 6.96875,
      "learning_rate": 2.4353154409148637e-05,
      "loss": 7.2923,
      "step": 502
    },
    {
      "epoch": 0.81,
      "grad_norm": 6.625,
      "learning_rate": 2.3959403439996907e-05,
      "loss": 7.3927,
      "step": 503
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.53125,
      "learning_rate": 2.356842794541516e-05,
      "loss": 7.0782,
      "step": 504
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.78125,
      "learning_rate": 2.318024219597196e-05,
      "loss": 7.1464,
      "step": 505
    },
    {
      "epoch": 0.82,
      "grad_norm": 55.75,
      "learning_rate": 2.2794860360410342e-05,
      "loss": 6.8995,
      "step": 506
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.4375,
      "learning_rate": 2.241229650513077e-05,
      "loss": 7.1063,
      "step": 507
    },
    {
      "epoch": 0.82,
      "grad_norm": 6.375,
      "learning_rate": 2.2032564593677774e-05,
      "loss": 7.1596,
      "step": 508
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.6875,
      "learning_rate": 2.165567848623009e-05,
      "loss": 7.2437,
      "step": 509
    },
    {
      "epoch": 0.82,
      "grad_norm": 4.53125,
      "learning_rate": 2.1281651939094992e-05,
      "loss": 7.1674,
      "step": 510
    },
    {
      "epoch": 0.82,
      "grad_norm": 3.984375,
      "learning_rate": 2.0910498604205986e-05,
      "loss": 7.1779,
      "step": 511
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.3125,
      "learning_rate": 2.0542232028624586e-05,
      "loss": 7.0682,
      "step": 512
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.8125,
      "learning_rate": 2.0176865654045974e-05,
      "loss": 7.4579,
      "step": 513
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.53125,
      "learning_rate": 1.981441281630816e-05,
      "loss": 7.3214,
      "step": 514
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.6875,
      "learning_rate": 1.94548867449054e-05,
      "loss": 7.0165,
      "step": 515
    },
    {
      "epoch": 0.83,
      "grad_norm": 28.75,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 7.1816,
      "step": 516
    },
    {
      "epoch": 0.83,
      "grad_norm": 5.78125,
      "learning_rate": 1.8744667284469575e-05,
      "loss": 7.3203,
      "step": 517
    },
    {
      "epoch": 0.84,
      "grad_norm": 4.9375,
      "learning_rate": 1.8393999818379525e-05,
      "loss": 7.1352,
      "step": 518
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.4375,
      "learning_rate": 1.804631096356435e-05,
      "loss": 7.1253,
      "step": 519
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.65625,
      "learning_rate": 1.7701613410634365e-05,
      "loss": 6.9686,
      "step": 520
    },
    {
      "epoch": 0.84,
      "grad_norm": 38.5,
      "learning_rate": 1.735991974101756e-05,
      "loss": 7.4466,
      "step": 521
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.125,
      "learning_rate": 1.7021242426500493e-05,
      "loss": 7.0422,
      "step": 522
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.0625,
      "learning_rate": 1.6685593828773095e-05,
      "loss": 7.2777,
      "step": 523
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.4375,
      "learning_rate": 1.6352986198977325e-05,
      "loss": 7.1547,
      "step": 524
    },
    {
      "epoch": 0.85,
      "grad_norm": 6.3125,
      "learning_rate": 1.6023431677260214e-05,
      "loss": 6.975,
      "step": 525
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.5,
      "learning_rate": 1.5696942292330576e-05,
      "loss": 7.4417,
      "step": 526
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.625,
      "learning_rate": 1.5373529961019974e-05,
      "loss": 7.2602,
      "step": 527
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.78125,
      "learning_rate": 1.5053206487847914e-05,
      "loss": 7.2853,
      "step": 528
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.4375,
      "learning_rate": 1.4735983564590783e-05,
      "loss": 7.3027,
      "step": 529
    },
    {
      "epoch": 0.85,
      "grad_norm": 183.0,
      "learning_rate": 1.442187276985526e-05,
      "loss": 7.3557,
      "step": 530
    },
    {
      "epoch": 0.86,
      "grad_norm": 6.96875,
      "learning_rate": 1.4110885568655564e-05,
      "loss": 7.5958,
      "step": 531
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.65625,
      "learning_rate": 1.3803033311995072e-05,
      "loss": 6.8963,
      "step": 532
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.125,
      "learning_rate": 1.3498327236452013e-05,
      "loss": 7.6564,
      "step": 533
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.03125,
      "learning_rate": 1.3196778463769255e-05,
      "loss": 7.232,
      "step": 534
    },
    {
      "epoch": 0.86,
      "grad_norm": 3.984375,
      "learning_rate": 1.2898398000448443e-05,
      "loss": 7.1699,
      "step": 535
    },
    {
      "epoch": 0.86,
      "grad_norm": 6.75,
      "learning_rate": 1.260319673734821e-05,
      "loss": 7.1442,
      "step": 536
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.0625,
      "learning_rate": 1.2311185449286677e-05,
      "loss": 7.0898,
      "step": 537
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.75,
      "learning_rate": 1.2022374794648228e-05,
      "loss": 7.1733,
      "step": 538
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.3125,
      "learning_rate": 1.1736775314994342e-05,
      "loss": 6.8209,
      "step": 539
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.5625,
      "learning_rate": 1.1454397434679021e-05,
      "loss": 7.2807,
      "step": 540
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.8125,
      "learning_rate": 1.1175251460468117e-05,
      "loss": 7.2368,
      "step": 541
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.34375,
      "learning_rate": 1.0899347581163221e-05,
      "loss": 7.4665,
      "step": 542
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.9375,
      "learning_rate": 1.062669586722983e-05,
      "loss": 7.2071,
      "step": 543
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.03125,
      "learning_rate": 1.0357306270429624e-05,
      "loss": 7.148,
      "step": 544
    },
    {
      "epoch": 0.88,
      "grad_norm": 4.5,
      "learning_rate": 1.0091188623457415e-05,
      "loss": 7.1838,
      "step": 545
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.6875,
      "learning_rate": 9.828352639582072e-06,
      "loss": 7.2099,
      "step": 546
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.875,
      "learning_rate": 9.568807912292077e-06,
      "loss": 7.2912,
      "step": 547
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.875,
      "learning_rate": 9.31256391494546e-06,
      "loss": 7.1883,
      "step": 548
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.4375,
      "learning_rate": 9.05963000042378e-06,
      "loss": 7.5636,
      "step": 549
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.6875,
      "learning_rate": 8.810015400790994e-06,
      "loss": 7.1998,
      "step": 550
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.1875,
      "learning_rate": 8.563729226956319e-06,
      "loss": 7.3323,
      "step": 551
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.375,
      "learning_rate": 8.32078046834176e-06,
      "loss": 7.272,
      "step": 552
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.5625,
      "learning_rate": 8.081177992554013e-06,
      "loss": 7.573,
      "step": 553
    },
    {
      "epoch": 0.89,
      "grad_norm": 4.375,
      "learning_rate": 7.844930545060703e-06,
      "loss": 7.2687,
      "step": 554
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.15625,
      "learning_rate": 7.612046748871327e-06,
      "loss": 7.0265,
      "step": 555
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.875,
      "learning_rate": 7.382535104222366e-06,
      "loss": 7.5013,
      "step": 556
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.0625,
      "learning_rate": 7.156403988267069e-06,
      "loss": 7.6454,
      "step": 557
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.96875,
      "learning_rate": 6.9336616547697965e-06,
      "loss": 7.377,
      "step": 558
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.53125,
      "learning_rate": 6.714316233804574e-06,
      "loss": 7.2638,
      "step": 559
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.5625,
      "learning_rate": 6.498375731458528e-06,
      "loss": 7.3772,
      "step": 560
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.71875,
      "learning_rate": 6.28584802953951e-06,
      "loss": 7.4035,
      "step": 561
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.0,
      "learning_rate": 6.076740885288479e-06,
      "loss": 7.2213,
      "step": 562
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.5625,
      "learning_rate": 5.8710619310964445e-06,
      "loss": 7.3617,
      "step": 563
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.8125,
      "learning_rate": 5.668818674225685e-06,
      "loss": 7.4216,
      "step": 564
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.65625,
      "learning_rate": 5.470018496535967e-06,
      "loss": 7.2892,
      "step": 565
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.75,
      "learning_rate": 5.274668654214932e-06,
      "loss": 7.1227,
      "step": 566
    },
    {
      "epoch": 0.91,
      "grad_norm": 6.5,
      "learning_rate": 5.08277627751329e-06,
      "loss": 7.4076,
      "step": 567
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.65625,
      "learning_rate": 4.8943483704846475e-06,
      "loss": 7.0772,
      "step": 568
    },
    {
      "epoch": 0.92,
      "grad_norm": 7.1875,
      "learning_rate": 4.709391810729713e-06,
      "loss": 6.7843,
      "step": 569
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.40625,
      "learning_rate": 4.527913349145441e-06,
      "loss": 7.2041,
      "step": 570
    },
    {
      "epoch": 0.92,
      "grad_norm": 6.25,
      "learning_rate": 4.349919609678455e-06,
      "loss": 7.3247,
      "step": 571
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.75,
      "learning_rate": 4.175417089083378e-06,
      "loss": 7.1626,
      "step": 572
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.375,
      "learning_rate": 4.004412156685711e-06,
      "loss": 7.3347,
      "step": 573
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.6875,
      "learning_rate": 3.836911054149239e-06,
      "loss": 6.8028,
      "step": 574
    },
    {
      "epoch": 0.93,
      "grad_norm": 5.34375,
      "learning_rate": 3.6729198952483724e-06,
      "loss": 7.1392,
      "step": 575
    },
    {
      "epoch": 0.93,
      "grad_norm": 6.3125,
      "learning_rate": 3.512444665644865e-06,
      "loss": 7.2341,
      "step": 576
    },
    {
      "epoch": 0.93,
      "grad_norm": 7.6875,
      "learning_rate": 3.355491222669371e-06,
      "loss": 7.2214,
      "step": 577
    },
    {
      "epoch": 0.93,
      "grad_norm": 6.3125,
      "learning_rate": 3.202065295107726e-06,
      "loss": 7.2928,
      "step": 578
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.5,
      "learning_rate": 3.052172482991711e-06,
      "loss": 7.2009,
      "step": 579
    },
    {
      "epoch": 0.94,
      "grad_norm": 4.125,
      "learning_rate": 2.905818257394799e-06,
      "loss": 7.1589,
      "step": 580
    },
    {
      "epoch": 0.94,
      "grad_norm": 4.09375,
      "learning_rate": 2.7630079602323442e-06,
      "loss": 7.1614,
      "step": 581
    },
    {
      "epoch": 0.94,
      "grad_norm": 5.59375,
      "learning_rate": 2.6237468040666512e-06,
      "loss": 7.1077,
      "step": 582
    },
    {
      "epoch": 0.94,
      "grad_norm": 5.875,
      "learning_rate": 2.4880398719167586e-06,
      "loss": 7.1149,
      "step": 583
    },
    {
      "epoch": 0.94,
      "grad_norm": 5.0,
      "learning_rate": 2.3558921170727888e-06,
      "loss": 7.4501,
      "step": 584
    },
    {
      "epoch": 0.94,
      "grad_norm": 6.75,
      "learning_rate": 2.2273083629153147e-06,
      "loss": 7.0099,
      "step": 585
    },
    {
      "epoch": 0.95,
      "grad_norm": 5.6875,
      "learning_rate": 2.1022933027391555e-06,
      "loss": 7.5229,
      "step": 586
    },
    {
      "epoch": 0.95,
      "grad_norm": 5.3125,
      "learning_rate": 1.9808514995821593e-06,
      "loss": 7.1471,
      "step": 587
    },
    {
      "epoch": 0.95,
      "grad_norm": 7.5,
      "learning_rate": 1.8629873860586566e-06,
      "loss": 7.0446,
      "step": 588
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.375,
      "learning_rate": 1.7487052641976032e-06,
      "loss": 7.0627,
      "step": 589
    },
    {
      "epoch": 0.95,
      "grad_norm": 4.9375,
      "learning_rate": 1.6380093052856483e-06,
      "loss": 7.0851,
      "step": 590
    },
    {
      "epoch": 0.95,
      "grad_norm": 6.34375,
      "learning_rate": 1.5309035497147684e-06,
      "loss": 7.5415,
      "step": 591
    },
    {
      "epoch": 0.95,
      "grad_norm": 6.0625,
      "learning_rate": 1.4273919068349184e-06,
      "loss": 7.3371,
      "step": 592
    },
    {
      "epoch": 0.96,
      "grad_norm": 5.6875,
      "learning_rate": 1.3274781548112458e-06,
      "loss": 7.2418,
      "step": 593
    },
    {
      "epoch": 0.96,
      "grad_norm": 5.71875,
      "learning_rate": 1.231165940486234e-06,
      "loss": 7.7067,
      "step": 594
    },
    {
      "epoch": 0.96,
      "grad_norm": 7.40625,
      "learning_rate": 1.1384587792465872e-06,
      "loss": 7.2579,
      "step": 595
    },
    {
      "epoch": 0.96,
      "grad_norm": 7.03125,
      "learning_rate": 1.0493600548948878e-06,
      "loss": 7.477,
      "step": 596
    },
    {
      "epoch": 0.96,
      "grad_norm": 7.0625,
      "learning_rate": 9.638730195261625e-07,
      "loss": 7.0403,
      "step": 597
    },
    {
      "epoch": 0.96,
      "grad_norm": 5.09375,
      "learning_rate": 8.820007934090879e-07,
      "loss": 7.3635,
      "step": 598
    },
    {
      "epoch": 0.97,
      "grad_norm": 7.96875,
      "learning_rate": 8.037463648721488e-07,
      "loss": 7.5187,
      "step": 599
    },
    {
      "epoch": 0.97,
      "grad_norm": 362.0,
      "learning_rate": 7.291125901946027e-07,
      "loss": 7.179,
      "step": 600
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.8125,
      "learning_rate": 6.581021935021304e-07,
      "loss": 7.0963,
      "step": 601
    },
    {
      "epoch": 0.97,
      "grad_norm": 5.78125,
      "learning_rate": 5.907177666674812e-07,
      "loss": 7.2989,
      "step": 602
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.53125,
      "learning_rate": 5.269617692158613e-07,
      "loss": 7.1139,
      "step": 603
    },
    {
      "epoch": 0.97,
      "grad_norm": 7.21875,
      "learning_rate": 4.668365282351372e-07,
      "loss": 7.4399,
      "step": 604
    },
    {
      "epoch": 0.98,
      "grad_norm": 5.25,
      "learning_rate": 4.103442382909051e-07,
      "loss": 7.2847,
      "step": 605
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.625,
      "learning_rate": 3.5748696134639825e-07,
      "loss": 7.1542,
      "step": 606
    },
    {
      "epoch": 0.98,
      "grad_norm": 5.90625,
      "learning_rate": 3.0826662668720364e-07,
      "loss": 7.3664,
      "step": 607
    },
    {
      "epoch": 0.98,
      "grad_norm": 6.625,
      "learning_rate": 2.6268503085089547e-07,
      "loss": 7.3751,
      "step": 608
    },
    {
      "epoch": 0.98,
      "grad_norm": 13.875,
      "learning_rate": 2.2074383756137686e-07,
      "loss": 7.2611,
      "step": 609
    },
    {
      "epoch": 0.98,
      "grad_norm": 7.03125,
      "learning_rate": 1.824445776682504e-07,
      "loss": 7.6417,
      "step": 610
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.34375,
      "learning_rate": 1.477886490908742e-07,
      "loss": 7.269,
      "step": 611
    },
    {
      "epoch": 0.99,
      "grad_norm": 4.625,
      "learning_rate": 1.1677731676733584e-07,
      "loss": 7.2,
      "step": 612
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.3125,
      "learning_rate": 8.941171260835601e-08,
      "loss": 7.2481,
      "step": 613
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.34375,
      "learning_rate": 6.569283545587724e-08,
      "loss": 7.1505,
      "step": 614
    },
    {
      "epoch": 0.99,
      "grad_norm": 4.875,
      "learning_rate": 4.562155104665955e-08,
      "loss": 7.1732,
      "step": 615
    },
    {
      "epoch": 0.99,
      "grad_norm": 4.28125,
      "learning_rate": 2.9198591980705848e-08,
      "loss": 7.1312,
      "step": 616
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.25,
      "learning_rate": 1.642455769444995e-08,
      "loss": 7.2225,
      "step": 617
    },
    {
      "epoch": 1.0,
      "grad_norm": 4.875,
      "learning_rate": 7.2999144389296335e-09,
      "loss": 7.3122,
      "step": 618
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.34375,
      "learning_rate": 1.8249952627669154e-09,
      "loss": 7.2489,
      "step": 619
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.9375,
      "learning_rate": 0.0,
      "loss": 6.8919,
      "step": 620
    },
    {
      "epoch": 1.0,
      "eval_loss": 7.301526069641113,
      "eval_runtime": 1.7862,
      "eval_samples_per_second": 19.035,
      "eval_steps_per_second": 19.035,
      "step": 620
    }
  ],
  "logging_steps": 1,
  "max_steps": 620,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 7881494668247040.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}